vvEPA
United States
Environmental Protection
Agency
Office of Solid Waste
and Emergency Response
Washington DC 20460
July 1986
Solid Watte
Criteria for Identifying
Areas of Vulnerable
Hydrogeology Under the
Resource Conservation
and Recovery Act
Appendix A
Technical Methods for
Evaluating
Hydrogeologic
Parameters
Interim Final
-------
0^,; .'LICY DIRECTIVE NO.
9472 -00-2A
TABLE OF CONTENTS
Appendix A
Table of Contents i
List of Figures iii
List of Tables iii
1.0 Methods to Characterize Hydraulic Conductivity.... 1-1
1.1 Well Construction Considerations 1-3
1.1.1 Wells Requiring Screens 1-4
1.1.2 Wells Not Requiring Screens 1-5
1. 2 Well Development 1-6
1.3 Data Interpretation and Test Selection
Considerations 1-6
1.4 Single Well Tests 1-8
1.4.1 Methods for Moderately Permeable Formations
Under Confined Conditions 1-8
1.4.2 Methods for Extremely Tight Formations
Under Confined Conditions 1-15
1.4.3 Methods for Moderately Permeable Materials
Under Unconfined Conditions 1-17
1. 5 Multiple Well Tests 1-24
1. 6 Fractured Media 1-24
2.0 Determination of- Effective Porosity 2-1
2.1 Specific Yield in the Unconfined Aquifer 2-1
2 . 2 Confined Aquifers 2-5
2.3 Tabulated Values of Effective Porosity 2-5
3.0 Determination of Hydraulic Gradient 3-1
4.0 References 4-1
4.1 References for Section 1.0 4-1
-------
OSWER POLICY DIRECTIVE j:-j
9472•00-2A
4.2 References for Section 2.0 7. ..74-2
4. 3 References for Section 3.0 ,. 4-2
ii
-------
LIST OF FIGURES
1.4-1 Geometry and Variable Definitions for Slug
Tests in Confined Aquifers 1-10
1.4-2 Schematic Diagram for Pressurized Slug
Test Method in Consolidated and
Unconsolidated Materials 1-17
1.4-3 Variable Definitions for Slug Tests in
Unconfined Materials 1-20
1.4-4 Curves Defining Coefficients A, B, and
C in Equations 13 and 14 as a function
of the Ratio L/rw 1-23
2.0-1 Concept of Specific Yield veiwed in terms of
the Unsaturated Moisture Profile Above the
Water Table 2-2
2.0-2 Relationship between Specific Yield and
Grain Size 2-4
3.0-1 Diagram of How Temporal Changes in Hydraulic
Gradient Affect the Flow Paths of Ground-Water.3-3
3.0-2 Illustration of Mounding in the Unsaturated
Zone 3-4
3.0-3 Diagram of Mounding in the Saturated Zone 3-6
LIST OF TABLES
1.4-1 Values of the Function F (<<,£) for use in
the conventional and pressurized slug tests.... 1-12
iii
-------
FIELD METHODS TO DETERMINE
HYDRAULIC CONDUCTIVITY
SECTION 1.0
This section discusses methods available for the determination
of fluid conductivity under field conditions. As most of these
tests will use water as the testing fluid, either natural for-
mation water or water added to a borehole or piezometer, the
term hydraulic conductivity will be used for the remainder of
this section. Before testing for hydraulic condu&tivity, the
first step is to characterize the geology of the site. This
initial characterization will aid i'n the design of well locations
The investigator should consult the RCRA Ground-Water Monitoring
Technical Enforcement Guidance Document, U.S.EPA (Draft) 1985
(TEGD) for information on site characterization.
The owner/operator must identify areas of high and low hydraulic
conductivity (K) within each formation, because such variations
in K can create irregularities in ground water flow paths.
Areas of high hydraulic conductivity represent areas of greater
ground water flow and, if contaminants are present, zones of
potential migration.
Hydraulic conductivity measurements must define both vertical
and horizontal hydraulic conductivity across an owner/operator's
regulated site. In assessing the completeness of an owner/
operator's hydraulic conductivity measurements, the reviewer
1-1
-------
should also consider results trom the boring program used to
characterize the site geology. Zones of high permeability or
fractures identified from drilling logs should have been con-
sidered in the determination of hydraulic conductivity.
Additionally, information from coring logs can be used to
refine the data generated by slug or pump tests.
The location of wells, selection of screened intervals, and the
appropriate tests depend upon the specific site under invest-
igation. The person responsible for such selections should be
a qualified hydrogeologist or geotechnical engineer experienced
in the application of established principles of contaminant
hydrogeology and ground water hydraulics.
To estimate hydraulic conductivity, single well methods are
the preferred method since they do not average the effects
of geologic heterogeneities. In addition, the methods for
field determination of hydraulic conductivity are restricted
to well or piezometer tests applicable below existing
water tables. Determination of travel times of leachate and
dissolved solutes above the water table usually require the
application of unsaturated flow theory. Methods for calculating
travel times in unsaturated zones are detailed in Appendix C.
Standard reference texts on ground water hydraulics and con-
taminant hydrogeology that should be consulted include: Bear
(1972), Bouwer (1978), Freeze and Cherry (1979), Stallman (1971),
and Walton (1970) .
1-2
-------
D(S£cm.
1.1 WELL-CONSTRUCTION CONSIDERATIONS
The purpose of using properly constructed wells for hydraulic
conductivity testing is to assure that test results reflect
conditions in the geologic interval being tested, rather than
the conditions caused by well construction. In all cases,
diagrams showing all details of the actual well or borehole
constructed for the test should be made, as well as detailed
narrative description.
Detailed discussions of well installation, construction, and
r'-
development methods are given by Bouwer (1978), pages 160-180,
Acker (1974), Johnson (1975), and the TEGD, Chapter 3.
The TEGD should be consulted for information concerning well
drilling, construction, and development considerations. However,
the TEGD is concerned primarily with ground water quality mon-
itoring wells. Boreholes suitable for measuring hydraulic
conductivity may not necessarily also be suitable for ground
water quality monitoring, due to considerations unique to mon-
itoring (such as the need for inert casing, possible need for
pumps to be installed, etc.). The TEGD discusses these differ-
ences in more detail.
1.1.1 Wells Requiring Well Screens
Well screens placed opposite the interval to be tested should
be constructed of materials that are compatible with the fluids
to be encountered. The screen slot size should be determined to
minimize the inflow of fine-grained material to the well during
development and testing. In addition, hydraulic conductivity
1-3
-------
information generally provides average values for the entire area
across a well screen. For more depth discrete information, well
screens will have to be shorter. If the average hydraulic con-
ductivity for a formation is required, entire formations may have
to be screened. Bouwer (1978), and Johnson (1975) give a summary of
guidelines for sizing well screens.
The annulus between the well screen and the borehole should be
filled with an artificial gravel pack or sand filter. Guide-
lines for sizing these materials are given by Johrfson (1975).
For very coarse materials, it may be acceptable to allow the
materials from the tes'ted zone to collapse around the screen
forming a natural gravel pack.
The screened interval should be isoloated from overlying and
underlying zones by materials of low hydraulic conductivity.
Generally, a short bentonite plug is placed on top of the
material surrounding the screen, and cement grout is placed in
the borehole to the nextihigher screened interval (in the case
of multiple screen wells), or to the land surface for single
screen wells.
Although considerations for ground water quality sampling may
dictate different minimum casing and screen diameters, the
recommended guideline for hydraulic conductivity characterization
studies is that wells to be tested by pumping, bailing, or injection
in coarse-grained materials should be at least 4 inches inside
diameter. Wells to be used for testing materials of low hydraulic
1-4
-------
conductivity by sudden removal or injection of a known volume
of fluid should be constructed with as small a casing diameter
as possible to maximize measurement resolution of fluid level
changes. Casing sizes of 1.25 to 1.50 inches usually allow
this resolution while enabling the efficient sudden withdrawal
of water for these tests.
1.1.2 Wells Not Requiring Well Screens
If the zone to be tested is sufficiently indurated that a well
screen and casing is not required to prevent caving/ then it is
rr- ' ' .
preferable to use a borehole open to the zone to be tested.
These materials generally are those having low to extremely
low hydraulic conductivities. Consolidated rocks having high
conductivity because of the presence of fractures and solution
openings may also be completed without the use of a screen and
gravel pack. Uncased wells may penetrate several zones for
which hydraulic conductivity test are to be run. In these
cases, the zones of interest may be isolated by the use of
inflatable packers.
1.2 WELL DEVELOPMENT
For wells that are constructed with well screens and gravel
packs, and for all wells in which drilling fluids have been
used that may have penetrated the materials to be tested, ade-
quate development of the well is required to remove these
fluids and to remove the fine-grained materials from the zone
around the well screen. Development is carried out by methods
such as intermittent pumping, jetting with water, surging, and
1-5
-------
bailing. Adequate development is required to assure maximum
communication between fluids in the borehole and the zone to be
tested. Results from test run in wells that are inadequately
developed will include an error caused by-loss of fluid poten-
tial across the undeveloped zone, and computed hydraulic con-
ductivities will be lower than the actual value. Bouwer (1978),
Johnson (1975), and the TEGD give further details on well
development including methods to determine' when adequate
development has occurred.
1.3 DATA INTERPRETATION AND TEST SELECTION CONSIDERATIONS
Hydraulic conductivity may be determined in wells that are
either cased or uncased as described in Section l.l. The tests
all involve disturbing the existing fluid potential in the
tested zone by withdrawal from or injection of fluid into a
well either as a slug over an extremely short period of time,
or by continuous withdrawal or injection of fluid. The
hydraulic conductivity is determined by measuring the response
of the water level or pressure in the well as a function of
time since the start of the test. Many excellent references
are available that give the derivation and use of the methods
that are outlined below, including Bouwer (1978), Walton
(1969), and Lohman (1972).
1-6
-------
The selection of a particular test method and data analysis
technique requires the consideration of the purposes of the
test, and the geologic framework in which the test is to be
run. Knowledge of the stratigraphic relationships of the zone
to be tested and both overlying and underlying materials should
always be used to select appropriate test design and data
interpretation methods.
The equations given for' all computational methodsvgiven here
and in the above references are based on idealized .models
comprising layers of materials of different hydraulic conduc-
tivities. The water-level response caused by disturbing the
system by the addition or removal of water can be similar for
quite different systems. For example, the response of a water-
table aquifer and a leaky, .confined aquifer to pumping can be
very similar. Consequently, it is not considered acceptable
practice to obtain data from a hydraulic conductivity test and
interpret the type of hydraulic system present without sup-
porting geologic evidence.
The well test methods are discussed under the following two
categories: 1) methods applicable to coarse-grained materials
and tight to extremely*tight materials under confined conditions;
and 2) methods applicable to unconfined materials of moderate
1-7
-------
permeability. The single well tests integrate the effects of
heterogeneity and anisotropy. The effects of boundaries such
as streams or less permeable materials usually are not detec-
table with these methods because of the small portion of the
geologic unit that is tested.
1.4 SINGLE WELL TESTS
The tests for determining hydraulic conductivity with a single
well are discussed below based on methods for confined and
unconfined conditions. The methods are usually called slug
tests because the test involves removing a slug of water
instantaneously from a well and measuring the recovery of water
in the well. The method was first developed by Hvorslev (1951),
whose analysis did not consider the effect of fluid stored in
the well. Cooper and others (1967) developed a method that
considers well bore storage. However, their method only
applies to wells that are open to the entire zone to be tested
and that tap confined aquifers. Because of the rapid water-
level response in coarse materials, the tests are generally
limited to zones with a transmissivity of less than about
70 cmVsec (Lohman, 1972). The method has been extended to
allow testing of extremely tight formations by Bredehoeft and
Papadopulos (1980). Bouwer and Rice (1976) developed a method
for analyzing slug tests for unconfined aquifers.
1 .4.1 Method for Moderately Permeable Formations Under
Confined Conditions
1.4.1.1 Applicability. This method is applicable for testing
zones to which the entire zone is open to the well screen or
1-8
-------
open borehole. The method usually is used in materials of
moderate hydraulic conductivity which allow measurement of
water-level response over a period of an hour to a few days.
More permeable zones can be tested with rapid response water-
level recording equipment. The method assumes that the tested
zone is uniform in all radial directions from the test well.
Figure 1.4-1 illustrates the test geometry for this method.
1.4.1.2 Procedures. The slug test is run by utilizing some
method of removing a known volume of water from the well bore
in a very short time period and measuring the recovery of the
water level in the well. The procedures are the same for both
unconfined and confined aquifers. Water is most effectively
removed by using a bailer that has been allowed to fill and
stand in the well for a sufficiently long period of time so
that any water-level disturbance caused by the insertion of the
bailer will have reached equilibrium, in permeable materials,
this recovery time may be as little as a few minutes. An
alternate method of effecting a sudden change in water level is
the withdrawal of a weighted float. The volume of water
displaced can be computed using the known submersed volume of
the float and Archimedes' principle (Lohman, 1972).
Water-level changes are recorded using either a pressure trans-
ducer and a strip chart recorder, a weighted steel tape, or an
electric water-level probe. For testing permeable materials
that approach or exceed 70 cm2/sec, a rapid-response transducer/
recorder system is usually used because essentially full reco-
very may occur in a few minutes. Because the rate of water-
level response decays with time, water-level or pressure
1-9
-------
WELL CASING
WELL SCREEN
/ CONFINING LAYER
Figure 1.4-1—Geometry and variable definition for
slug tests in confined aquifers.
1-10
-------
changes should be taken at increments that are approximately
equally spaced in the logarithm of the time since fluid
withdrawal. The test should be continued until the water level
in the well has recovered to at least 85 percent of the initial
pre-test value.
1.4.1.3 Calculations. Calculations for determining hydraulic
conductivity for moderately permeable formations under confined
conditions can be made using the following:
a. Determine the transmissivity of the tested zone by
plotting the ratio h/ho on an arithmetic scale against
time since removal of water Ct) on a logarithmic scale.
The observed fluid potential in the well during the test
as measured by water level or pressure is h, and ho is
the fluid potential before the instant of fluid
withdrawal. The data plot is superimposed on type
curves, such as those given by Lohman (1972), Plate 2
or plotted from Table 1.4-1 with the h/ho and time axes
coincident. The data plot is moved horizontally until
the data fits one of the type curves. A value of time
on the data plot corresponding to a dimensionless time
(0) on the type curve plot is chosen, and the transmis-
sivity is computed from the following:
T = 6rc2> (1)
t
where rc is the radius of the casing (Lohman, 1972, p. 29)
1-11
-------
TABLE 1.4-1
Values of the function F(a,6) for use in
the conventional and pressurized slug tests.
Source: Papadopulos et.al. (1973)
rt/r,'
0.001
0.002
0.004
0.006
0.008
0.01
0.02
0.04
0.06
0.08
0.1
0.2
0.4
0.6
0.8
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
20.0
30.0
40.0
50.0
60.0
80.0
100.0
200.0
a - 10-«
0.9994
0.99S9
0.99SO
0.9972
0.9964
0.9956
0.9919 .
0.9S48
0.9782
0.9718
0 9655
0.9361
0.8828
0.8345
0.7901
0.7489
0.5800
0.4554
0.3613
0.2S93
0.2337
0.1903
0.1562
0.1292
0.1078
0.02720
0.01286
0.008337
0.006209
0.004961
0.003547
0.002763
0.001313
o - 10"
0.9996
0.9992
0.9985
0.9978
0.9971
0.9965
0.9934
0.9S75
0.9819
0.9765
0.9712
0.9459
0.8995
0.8569
0.8173
0.7801
0.6235
0.5033
0.4093
0.3351
0.2759
0.2285
0.1903
0.1594
0.1343
0.03343
0.01448
O.OOSS98
0.006470
0.005111
0.003617
0.002803
0.001322
a - 10-«
0.9996
0.9993
0.9987
0.9982
0.9976
0.9971
0.9944
,0.9894
0.9846
0.9799
0.9753
0.9532
0.9122
0.8741
0'.83S3
0.8045
0.6591
0.5442
0.4517
0.3768
0.3157
0.2655
0 2243
0.1902
0.1620
0.04129
0.01667
0.009637
0.006789
• 0.005283
0.003691
0.002845
0.001330
a - 10-»
0.9997
0.9994
0.9989
0.9984
0.9980
0.9975
0.9952
0.9908
0 9S66
0.9824
0.9784
0.9587
0.9220
0.8875
0.8550
0.8240
0.6889
0.5792
0.4891
0.4146
0.3525
0.3007
0.2573
0.2208
0.1900
0.05071
0.01956
0.01062
0.007192
0.005487
0.003773
0.002890
0.001339
a - 10-"
0.9997
0.9995
0.9991
0.9986
0.9982
0.9978
0.9958
0.9919
0.9881
0.9844
0.9807
0.9631
0.9298
0.8984
0.8686
0.8401
0.7139
0.6096
0.5222
0.4487
0.3865
0.3337
0.2888
0.2505
0.2178
0.06149
0.02320
0.01190
0.007709
0.005735
0.003S63
0.002938
0.001348
1-12
-------
TABLE 1.4-1, continued
Extended values of F(ct,S) for use in slug tests.
Source: Bredehoeft and Papadooulos (1980).
p
c.oocooi
c.oocoo?
G.onc on*
c.oocoot
c.orc OOP,
O.orcoi
C.OPCO?
O.nnr o*
o.oocr*
o.opr PB
O.OOCI
c.oncr1
O.OOC*
c.ocr*
o.onrh
C.OPl
C.OCi
c.op*
c.cnt
O.cnt
C.Ol
0.02
0.0*
n.o*
C.OP
0.1
0.2
0.4
O.C
O.f>
1.
t «
«.
t.
e.
1C.
?c.
4C.
»-C.
MC.
inc.
?no.
• «C.
60C.
enc.
100C.
o = 0.1
P.°C93
P. 9990
u.sKbf
0.99«?
1.0900
O.V977
0.996S
J.'i'SSS
n.944«
0.9<)lb
.i.oq?"!
P.-J69C
O.'?0^'-
O.^PZJ
T.*79*
O.T765
n.9«,70
0.«»%3P
0.9*17
0.93??
0.9?3t
0.890*
O.fl*?l
o.en*f
n.773*
0.7*59
0.h41P
0.509?
(•..*? ?7
0.159(>
0.3117
0. 1 Tbf-
1.nP7tl
0.065?7
0.0?Q*3
n.(n065
0.01 41l*
o.oneteo
n. OP4167
O.CC3?*?
0.002J77
3.nni?7]
r.noo'-"*07
o.ooo*!-'''
C.000?l«0
n.ocn?5lo
0 = 0.2
O.s-*«in
0.99ff.
a.99tn
C.9-J7C,
0.9971
0.9966
0.99':?.
0.99if
0.99J?
0.99C9
n.9-<;o
0.98?7
0.97S7
0.97SJ
0.9713
C.9-J79
U.9S*6
0.93i7
0_9?1 1
0.90S9
O.fl9f2
O.HS42
P. 7980
0.7t»»,
0.7190
o.eeei
C.S77*
p.**;p
0.36*?
O.?h7?
0.2*,*P
0. Ibl9
0.17t9*
.C.O*S99
0.03tbe
0.0?t7C
0.013M
O.nn»c^e
C. 01*318
C.Oft}?!*
o.oo;<;E9
O.on]?t*
0.000ft?9?
O.OOC41PP
C.OOC3137
o.oocpsoe
0=0.5
0.996.
0.9977
r,.9°t.e
0.9V61
0.99*%
P. 99*9
0.99?9
a.9i-9o
0.9877
0.9BSP
n.9B»i
0.977*
0.9»>fl5
P.961S
O.-^S'
".950=
0.93P7
P. 9031
O.Sfc?^
n.86<>*
n.ssoe
0.79*7
0.721*
0.'>fr97
n.6£89
o.svm
U.*'99
0.356*
0.2Bfe*
0.2397
P.20M
0.120?
P.0b*?0
0.0*3 31
0.03?1?*
0.026"p
0.01?.cA
0.00617*
0.00*??9
0.003163
n.oo?s?ft
0.00175P
0.000*?72
0.000*177
0.0001131
0.0002SO*
(1 = 1
C.9977
0.9968
C.99SS
C.994S
C.9936
0.9979
T.9900
e.9e«,e
C.98P7
o.9nno
C.'»777
0,96P7
0.9%dO
O.o*f?>
0."3P5
0.93JS
0.904W
0.3606
O.«*19
0.8202
O.P.017
C.7336
0.6*89
0.5919
0.5*A6
O.S1?7
c.*oio
O.?902
O.J311
0.1931
0. 1663
C.0991?
0.05521
P. 03830
0.0?9^3
0.0?376
0.01719
O.OOM71
0.00*1 3?
P.00310R
0.00?*87
0.00)2*7
0.00062*2
9.000*1*2
0.00n31?3
0.n002«Q9
0=2
P.996P
0.99S5
0.^93^
0.9)
0.001Z3P
0.000*195
0.000*1* 1
0.00031 10
0.0002*90
U=5
0.99*8
0.9927
0.9P.98
0.9876
0.9857
O.VCU1
0.9776
0.9687
0.9619
0.9562
0.95 1-2
0.9321
0.9061
O.BP69
O.B71 1
0.8576
0.6075
0.7*39
0.7Q01
0.6662
0.636*
0.5*50
0,4*5*
0.3P7?
0.3*69
0.3168
0.2313
0.1612
0.1280
0.1077
0.09375
0.059*0
0.03621
0.02663 '
0.0212S
0.0177*
0.0099*3
0.005395
O.f'03726
0.002853
0.002313
0.00119*
0.0006085
0.000*087
0.0003078
0.0002*69
0=/10
0.9923
0.989*
0.9853
0.982?
0.9796
0.9773
0.9683
0.9556
0.9*6*
0.9387
0.931P
0.9059
0.8711
0.8*58
0.8253
0.8079
0.7*50
0.666*
0.6178
0.5797
0.5*92
O.»517
0.3556
0.3030
0.2682
0.2*28
0.17*0
0.1207
0.09616
0.0613*
0.07120
0.0*620
0.02908
0.02185
0.01771
0.01*99
0.006716
0.00*696
0.003**5
0.002668
0.002161
0.001 1*9
fl.0005***
0.000*016
0.0003035
0.0002**0
1-13
-------
The type curves plotted using data in Table 1.4-1 are not
to be confused with those commonly referred to as 'Theis
Curves' which are used for pumping tests in confined
aquifers (Lohman, 1972). The type curve method is a
general technique of detemining aquifer parameters when
the solution to the descriptive flow equation involves
more than one unknown parameter. Although both the
storage coefficient and transmissivity of the tested
interval can be determined with the type curve method
for slug tests, determination of storage coefficients is
beyond the scope of this report. See Section 1.4.1.4
for further discussion of the storage coefficient.
If the data in Table 1.4-1 are used, a type curve for
each value of a is prepared by plotting F(a,6) on the
arithmetic scale and dimensionless time (B) on the
logarithmic scale of semi-log paper.
b. Determine the hydraulic conductivity by dividing the
transmissivity by the thickness of the tested zone.
1.4.1.4 Sources of Error. The errors that can arise in
conducting slug tests can be of three types: those resulting
from the well or borehole construction, measurement errors,
and data analysis error.
Well construction and development errors. This method assumes
that the entire thickness of the zone of interest is open to
1-14
-------
OSWfER POLICY DIRECTS ...
9472-00-2A
the well screen or boreholes and that flow is principally radial.
%
If this is not the case, the computed hydraulic conductivity
may be too high. If the well is not properly developed, the
computed conductivity will be too low.
/
Measurement errors can result from determining or recording the
fluid level in the borehole and the time of measurement
incorrectly. Water levels should be measured to at least an
accuracy of 1 percent of the initial water-level change. For
moderately permeable materials, time should be measured with
an accuracy of fractions of minutes, and for more permeable
materials, the time should be measured in terms of seconds or
fractions of seconds. The latter may require the use of a
rapid-response, pressure transducer and recorder system.
Data analysis errors. The type curve procedure requires
matching the data to one of a family of type curves, described
by the parameter ex, which is a measure of the storage in the
well bore and aquifer. Papadopulos and others (1973) show that
an error of two orders of magnitude in the selection of oc would
result in an error of less than 30 percent in the value of
transmissivity determined. Assuming no error in determining
the thickness of the zone tested, this is equivalent to a 30
percent error in the hydraulic conductivity.
1.4.2 Methods For Extremely Tight Formations Under
Confined Conditions
1.4.2.1 Applicability. This test is applicable to materials
that have low to extremely low permeability'such as silts,
1-15
-------
clays, shales, and indurated lithologic units. The test has
been used to determine hydraulic conductivities of shales of as
low as 10~10 cm/sec.
1.4.2.2 Procedures. The test described by Bredehoeft and
Papadopulos (1980) and modified by Neuzil (1982) is conducted
by pressurizing suddenly a packed off zone in a portion of a
borehole or well. The test is conducted using a system such as
Shown in Figure 1.4-2. The system is filled with water to a level
assumed to be equal to the prevailing water level. This step
is required if sufficiently large times have not elapsed since
the drilling of the well to allow full recovery of water
levels. A pressure transducer and recorder are used to monitor
pressure changes in the system for a period prior to the test
to obtain pressure trends preceding the test. The system is
pressurized by addition of a known volume of water with a
high-pressure pump. The valve is shut and the pressure decay
is monitored. Neuzil1s modification uses two packers with a
pressure transducer below the bottom packer to measure the
pressure change in the cavity and one between the two packers
to monitor any pressure change caused by leakage around the
bottom packer.
1.4.2.3 Calculations. The modified slug test as developed by
Bredehoeft and Papadopulos (1980) considered compress'ive
storage of water in the borehole. These authors considered
that the volume of the packed-off borehole did not change
1-16
-------
Pressure Gage Q
System Filled
with Water !
|
;
, — Formation^j-- [•
:- -fl
System Filled
jwith Water
Open Hole
"'Packer-
3F-— —
3-
Interval to-
be Tested —
:_ J& _- :-
Pump
(a)
(b)
Figure 1.4-2 —
Schematic diagram for pressurized slug
test method in unconsolidated (a) and
consolidated (b) materials. Source:
Papadopulos and Bredehoeft, 1980.
1-17
-------
during the test and that all compressive storage resulted in
compression of water under the pressure pulse. Neuzil {1982)
demonstrated that under some test conditions this is .not a
valid assumption. The computational procedure is the same in
either case. Data is plotted in the same manner as for the
conventional slug test and type curves are used from either
Lohman (1972, Plate 2) or plotted from data given in Table 1.4-1
as described in Section 1.1.1.3. The value of time (t) and
dimensionless time, (0), are determined in the same manner as
for the conventional tests. If compression of water only is
considered, transmissivity is computed by replacing rc2 by the
quantity (Vw CwP gA) in Equation 2.
8(VW Cy pg/ir)2
(2)
where
Vw is the volume of water in the packed-off cavity, L^;
•GW is the compressibility of water, LT2n~l,
p is the density of water ML~3; and
g is the accleration of gravity, LT~2.
If the compressive storage is altered by changing the volume of
the packed-off cavity (V), then the combined compressibility of
the water and the expansion of the cavity (Co) is used. Co is
computed by measuring the volume of water injected during
1-18
-------
pressurization (AV) and the pressure change AP for the
pressurization:
C0 =
VAP
(Neuzil, 1982 / page 440). Use of Co requires an accurate method
of metering the volume of water injected and the volume of the
cavity.
1.4.2.4 Sources of Error. The types of errors in this method
t
are the same as those for the conventional slug test. Errors
may also arise by inaccurate determination of the cavity volume
and volume of water injected. An additional assumption that is
required for this method is that the hydraulic properties of
the interval tested remain constant throughout the test. This
assumption can best be satisfied by limiting the initial
pressure change to a value only sufficiently large enough to
be measured (Bredehoeft and Papadopulos, 1980).
1.4.3 Methods for Moderately Permeable. Materials Under
Unconfined Conditions
1.4.3.1 Applicability. This method is applicable to wells
that fully or partially penetrate the interval of interest
(Figure 1.4-3). The hydraulic conductivity determined will be
principally the value in the horizontal direction (Bouwer and
Rice, 1976).
1.4.3.2 Procedures. A general method for testing cased wells
that partly or fully penetrate aquifers that have a water table
1-19
-------
WELL CASING
WELLSEAI
Lw
7
L*
L_L
*l
vei
1
GRAVEL PACK
Lw
L*
k
Lw
STATIC WATER
LEVEL
IMPENETRABLE STRATUM ''
' *Wr»\ '^ 11 • (« . * *
<•> CASED WITH SCREEN
(b) CASED. NO SCREEN, NO
CAVITY ENLARGEMENT
(c) OPEN BOREHOLE
Figure 1.4-3 —Variable definitions for slug tests in
unconfined materials. Cased wells are
open at the bottom.
1-20
-------
as the upper boundary of the zone to be tested was .developed by
Bouwer and Rice (1976). The geometry and dimensions that are
required to be known for the method are shown in Figure 1.4-3. The
test is accomplished by effecting a sudden change in fluid
potential in the well by withdrawal of either a bailer or sub-
mersed float as discussed in Section 1.4.1.2. Water-level
changes can be monitored with either a pressure transducer and
recorder, a wetted steel tape, or an electric water-level
sounder. For highly permeable formations, a rapid-response
transducer and recorder system is required. The resolution of
the transducer should be about 0.01 m.
1.4.3.3 Calculations. The hydraulic conductivity is calcu-
lated using the following equation in the notation of this
report, taken from Bouwer and Rice (1976)
2 ~
rc In R/r ,_ ,_
K = __* ln ±2 , (3)
2 Let Y
where rc, rw, Le, t, Y, and K have been previously defined or
are defined in Figure 1.4-2(a). y i« the value of Y immediately
after withdrawal of the slug of water. The term R is an effec-
tive radius that is computed using the following equation given
by Bouwer and Rice (1976).
R 1 1.1 . A + B ln[(H0-Lw)/rw] ~1
ln «_ = t + } t (4)
rw In (Lw/rw (Le/rw)
1-21
-------
for wells that do not fully penetrate the aquifer, if the quan-
tity (H0-Lw)/rw) is larger than 6, a value of 6 should be used.
For wells that completely penetrate the aquifer, the .following
equation is used:
_1
In — » ( i^i + —£— ) , (5)
rw In (Lw/rw) Le/rw
(Bouwer, 1976). The values of the constants A, B, and C are
given by Figure 1.4-4 (Bouwer and Rice, 1976).
For both cases, straight-line portions of plots of the
logarithm of Y or YO/Y against time- should be used to determine
the slope, In Yo/Y
t
Additional methods for tests under unconfined conditions are
summarized by Bower (1976) on pages 117-122. These methods are
modifications of the cased-well method described above that
apply either to an uncased borehole or to a well or piezometer
in which the diameter of the casing and the borehole are the
same (Figures i.4-3(b) and (c).
1.4.3.4 Sources of Error. The method assumes that flow of
water from above is negligible. If this assumption cannot be
met, the conductivities may be in error. Sufficient flow from
the unsaturated zone by drainage would result in a high conduc-
tivity value. Errors caused by measuring water levels and
recording time are similar to those discussed in Sections
1.4.1.4 and 1.4.2.4.
1-22
-------
14
A
and
C
12
10
6
X I , I . I I .1 I J_l
.t.l UP
B
10
90 100
500 1000
5000
Figure 1.4-4 —Curves defining coefficients A, B,
and C in equations 13 and 14 as
a function of the ratio L/rw.
Source: Bower and Rice, 1976.
1-23
-------
1.5 MULTIPLE WELL TESTS
Hydraulic conductivity can also be determined by conventional
pumping tests in which water is continuously withdrawn or
injected using one well, and the water-level response is
measured over time in or near more observation wells. These
methods generally test larger portions of aquifers than the
single well test discussed in Section 1.4. For some cir-
cumstances these tests may be appropriate in obtaining data to
use in satisfying requirements of Part 264 Subpart F. However,
the large possibility for non-uniqueness in interpretation,
problems involved in pumping contaminated fluids, and the
expense of conducting such tests generally preclude their use
in problems of contaminant hydrogeology. The following
references give excellent discussions of the design and
interpretation of these tests: Lohman (1972), Stallman (1971),
and Walton (1970)..
1.6 FRACTURED MEDIA
Determining the hydraulic properties of fractured media is a
difficult process. Darcy's Law is not strictly applicable
to flow through fractures. However, it can be applied empirically
to large bodies of fractured rock if the fractures are spaced
closely enough so that the rock acts hydraulically in a manner
similar to granular porous media. For rocks with low fracture
density, analyzing flow in individual fissures may be necessary.
If these fissures are wide, flow may not be laminar and Darcy's
1-24
-------
Law does not apply.
The upper limit of Darcy's Law can be determined with the help
of the Reynold's number, Re, defined for porous media as
Re = /?vd
where j& ~ density
JU- - viscosity
v = specific discharge
d = a representative length dimension for the porous
medium (e.g., mean pore dimension, mean grain
diameter,' etc. )
It is widely used to distinguish between laminar flow occuring
at low velocities and turbulent flow at higher velocities.
Darcy's Law generally remains valid for Re values between 1 and
10. In wide fractures, Re can exceed this limit, making flow
analysis very difficult.
Furthermore, since fractured rock has low porosity, expansion
and contraction due to hydraulic and pressure head stress
affects the size of the fractures. Thus K becomes a function
of this stress, and is no longer constant.
Finally, determining fractured flow is difficult because the
orientation of fractures and density of their spacings in
different directions can cause anisotropy and trending heterogeneity
for K values. With all of these difficulties, a methodology
for estimating K in fractured media is beyond the scope of this
document. The investigator should consult Sowers (1981) for
guidance in analyzing flow through fractured media.
1-25
-------
DETERMINATION OF EFFECTIVE POROSITY
SECTION 2.0
Calculation of time of travel (TOT) of ground water to
evaluate the vulnerability of the saturated zone requires data
on effective porosity, ne/ with respect to flow through the
porous media. Effective porosity is usually less than total
porosity in that only the interconnected pore space through
which fluid flow occurs is of interest. Effective porosity
does not take into consideration the pore space in which fluid
is practically immobile. Effective 'porosity is less than total
porosity in fine textured medium where adhesion (i.e., the
attraction of molecules of fluid to adjacent solid surfaces of
the porous matrix) is important or when the porous matrix includes
a large portion of dead-end pores. (Bear, 1979)
2.1 Specific Yield in the Unconfined Aquifer
Ground-watereand drainage engineers often equate the term
effective porosity with the specific yield, Sy, when referring to
an unconfined aquifer where the water table is at depth equal to
several times the height of the capillary fringe. In these cases,
the term drainable porosity is also used.
Specific yield, Sy, is defined as the volume of water that
an unconfined aquifer releases from storage per unit decline in
the water table. As discussed by Freeze & Cherry (1979), the
idea of specific yield is best visualized with reference to the
saturated-unsaturated interaction it represents. Figure 2.0-1
2-1
-------
to
I
XT7I
Moisture content, Q
n
/77/////////////'
///////////////////////////////////////
Figure 2.0-1 concept of specific yield viewed in terms of the unsaturated moisture profiles
above the water table. (Freeze & Cherry, 1979)
-------
OSIVER POLICY DIRECTIVE KQ,
shows the water-table position and the vertical profile of moisture
fc
content vs. depth in the unsaturated zone at two times, ~ti~a~nd
t2- The cross hatched area represents the volume of water released
from storage in a volume of unit cross section. If the water-table
drop represents a unit decline, the cross hatched area represents
the specific yield, Sy.
The volume of water retained in an aquifer per unit area and
unit decline of the water table is called specific retention, Sr.
In a homogeneous isotropic soil with a sufficiently deep water
table, Sy + Sr = n when rt is the total porosity and specific
retention is identical to the field capacity of the soil. Under
such conditions, Figure 2.0-2 shows the relationship between
specific yield and specific retention for various soils. (Bear,
1979)
Laboratory values of specific yield usually differ from
those taken in the field due to the effects of
0 aquifer stratification,
0 slow drainage of materials above the water table,
0 air entrapment in the zone below the water table,
0 water table positions,
* the rate of change of the water table elevation.
Actual yields from field tests are reduced up to one-third of
laboratory values of Sy. Therefore, McWhorter and Sunada (1977)
differentiate between laboratory and field values by defining a
bulk parameter, apparent specific yield, Sya, that incorporates
these field influences. Because apparent specific yield depends
2-3
-------
Well-sorted material
Average material
60%
50%
40%
30%
20%
10%
0%
10~3
Clay
io-2
Silt
io-1 10°
Sand
IO1
Gravel
Median grain size (mm)
IO2
Cobbles
Figure 2.0-2 Relationship between specific yi$ld and grain size (from Conkllng et, al., 193f,
as modified by Davis and DeWiest, 1966) Bear, 1979
-------
on more than just the physical properties of porous solids,
values of apparent specific yield may bear little or no relation
to the specific yield. Field determination of apparent specific
yield as defined above can be made from analysis of aquifer test
data using the Theis solution or the Jacob method (when applicable)
as described by McWhorter & Sunada. Field determined values of
apparent specific yields are preferred over laboratory values
for determining the TOTiQO since they provide a more accurate
representation of the drainage porosity of the water-table aquifer.
2.2 Confined Aquifers
Time of travel calculations as discussed in this document
concern ground-water flow along a 100 foot flow line originating
at the base of the hazardous waste unit where ground-water move-
ment through saturated zones usually occurs under phreatic
(unconfined) conditions. However some flow lines may originate
or travel through a confined aquifer. Determination of
effective porosity for confined aquifers is rather different
than for an unconfined aquifer.
First, the capacity of confined aquifers to release water
from storage is markedly different from that of unconfined aquifers
The storage capacity of water-table aquifers, termed specific
yield, is representative of the effective porosity of the porous
media under homogeneous, isotropic conditions (Section 2.1).
As discussed by Freeze & Cherry, water released from storage in a
confined aquifer, termed storativity, is not derived from drainage
2-5
-------
of the voids as is the case in unconfirmed aquifers. Rather, it
is a function of the porosity of the aquifer material, the
secondary effects of water expansion and overburden stress, and
aquifer geometry. Therefore, storativity values for confined
aquifers should not be used as default values for effective
porosity.
Effective porosity for a confined aquifer may be best
approximated by laboratory tests run on samples taken from
the site, in order to bypass the secondary effects incorporated
into storativity values, yet still yield an effective porosity
that is as site-specific as possible.
2.3 Tabulated Values of Effective Porosity
Selection of tabulated values of effective porosity based on
soil texture classification (Table 3.2-2 in Guidance Criteria
for Identifying Areas of Vulnerable Hydrogeology : RCRA Statutory
Interpretive Guidance, June 1986) must consider site-specific
conditions when evaluating the vulnerability of the ground water
to waste constituents. The tabulated values are valid approximations
of effective porosity only in homogeneous geologic units composed
of relatively coarse materials in which the water table is at a
sufficient depth so as to eliminate any influence from the
capillary fringe. For example, recent studies indicate that the
water in some unweathered glacial tills in connate water
from the time of deposition (Barari and Hedges, 1985; Bradbury
et al., 1985). This suggests that ground-water flow in some
materials may be very slow or nonexistent. The tabulated
2-6
-------
values should be reduced up to one-third to incorporate the
influence of factors such as air entrapped near the water table,
stratification of materials above the water table, water table
position and the rate of change of the water table elevation
(McWhorter and Sunada, 1977).
For confined aquifers where storativity values are inaccurate,
and for tight formations where specific yield values are unobtain-
able from field tests, tabulated values may be used as guidelines
for estimating effective porosity.
2-7
-------
FACTORS AFFECTING CHANGES IN
HYDRAULIC GRADIENT
SECTION 3.0
Accurate determination of vertical and horizontal hydraulic
gradients is very important in characterizing a site's flow
regime. Overall characterization of hydraulic gradient (i) should
be done by using nested piezometers that are screened in the same
geologic unit, as discussed in Section 3.0 of the main text, in
Appendix B, and in the RCRA Technical Enforcement Guidance Document
U.S.EPA (Draft) 1985, (TEGD). Because methods for determining
hydraulic gradient are well-documented, this section will only
discuss factors that influence variations in (i) over time.
Seasonal, temporal, and artificially-induced variations in
ground-water flow and in corresponding hydraulic gradients
should be evaluated with respect to the resultant effect on TOT
calculation. The direction and rate of ground-water flow and
the hydraulic gradient, can be influenced by one or more of the
following external factors:
- well pumping
tidal processes and fluctuation
intermittent natural variation in river stage (e.g., bank
storage)
- artificial recharge
natural seasonal variations
construction activities or change in land use
Off-site or on-site well pumping may effect both the rate
and direction of ground-water flow in time-continuous or
3-1
-------
discontinuous patterns. Well water-level measurements must be
frequent enough to detect such water use patterns.
As discussed in the TEGD, natural processes such as
riverine, estuarine, or marine tidal movement may result in
variations of well water levels and/or ground-water quality.
Ground-water/stream flow interactions often influence the
hydraulic gradient of the ground water in the area adjacent to
the stream. An increase in river stage may induce flow into
stream banks creating a temporary reversal of ground-water flow
as shown in Figure 3.0-1. As the river stage declines, the
flow is reversed. Transitional patterns that may shift ground-
water flow directions must be evaluated with respect to the
effect on the vulnerability of the aquifer.
Short-term recharge patterns may affect ground-water flow
patterns that are markedly different from ground-water flow
patterns from seasonal averages. Selection of the hydraulic
gradient term in the time of travel calculation should incorporate
any variation in ground-water flow that may result from seasonal
fluctuations. Short-term fluctuations in the ground-water flow
pattern that result from dewatering due to construction, changes
in land use, or artificial recharge should also be evaluated in
the determination of the hydraulic gradient (see Figure 3.0-2).
The frequency of well water-level measurements from which
hydraulic gradients are determined will depend upon the nature
of the particular processes influencing ground-water flow at
the site. Well water-level measurements taken on a monthly
basis may be adequate to characterize natural seasonal variations.
3-2
-------
U)
I
TEMPORARY REVERSAL OF GROUND-WATER FLOW DUE TO
FLOODING OF A RIVER OR STREAM
Tempo>»ry
leveiial ol
9>ound«vaiei How .
m
Figure 3.0-1 Diagram of how temporal changes in hydraulic gradient affect the
flow paths of groundwater.
-------
u>
Zone of aeration
and treatment
Infiltration
Refuse)
Percolation through
unsaturated zone
Recharge mound
^^^
Original water table
Figure 3.0-2.
Illustration of how artificial recharge through an unsaturated zone from
a waste unit causes a recharge mound which changes the initial head in
the saturated zone.
-------
However, some daily water-level measurements may be required to
evaluate the influence of external factors such as well pumping,
dewatering due to construction, or artificial recharge. Riverine,
estuarine, or marine tidal influences may affect ground-water
flow patterns over a shorter time period making hourly or
continuous water level monitoring desirable. Continuous water
level monitoring should be performed until a pattern of general
water level fluctuation is determined for the site. Minimum
and maximum fluctuations of water level dependent upon site
conditions (i.e., external factors affecting water level and/or
natural factors such as tidal influences) should be carefully
considered.
Variations in hydraulic gradient determined from ground-
water level monitoring will result in a range of ground-water
flow velocities and possible variaton of flow direction. Flow
nets should be updated to reflect hydraulic gradient fluctuations
in order to evaluate a 'worst-case' 100-foot flow line. Time
of travel calculations should be made over the range of ground-
water velocities to account for the effect of seasonal and temporal
factors on the ground-water flow system. For example, Figure
3.0-3 illustrates that during a low water table, the 'worst-case'
100 foot flow line would travel down into the aquifer. However,
during a season with high precipitation, another principal flow
line would appear, moving horizontally out along a near surface
'basement-seepage' pathway. Further evaluation would be necessary
in this case to calculate which flow path would represent the
fastest TOT.
3-5
-------
Precipitation
\y Y Y Y Y
SEASONAL HIGH
WATER.TABLE
CO
SEASONAL LOW
WATER TABLE
^Groundwater zone
contaminated by leachate
Figure 3.0-3 Diagram of mounding in the saturated zone. Mounding within the waste unit causes
an increase in hydraulic gradient, providing a greater driving force for the
contaminant plume. Also note that changes in the seasonal precipitation chance '
the water table elevation, opening a new shallow migration path alone the more
permeable upper soil level.
-------
The initial head as a result of infiltration through a
landfill cover and the build up of leachate on the bottom liner
must be determined. The Hydrologic Evaluation of Land Performance
(HELP) Model can be used to estimate water movement across, into,
through, and out of landfills for a wide variety of landfill
designs. In the context of evaluating vulnerability as a
function of hydraulic gradient, the build up of leachate on or
beneath the landfill liner and its related impact on flow
velocity is of concern. Section 5 of the U.S. EPA Draft Background
Document entitled, Technical Support for Development and Analysis
of Hazardous Waste Disposal Regulations, March 1986, addresses
the use of the HELP model in estimating head at a landfill
liner. The HELP model expresses leachate head on the liner in
inches at designated time intervals. The calculation techniques
and inherent assumptions associated with the HELP model are
discussed in the referenced Draft Background document. The
major assumption impacting the accuracy of leachate head is that
of fully saturated conditions within the liner. This assumption
must be considered in determining hydraulic gradient for
calculating flow velocity.
3-7
-------
REFERENCES
SECTION 4.0
4.1 References for Section 1.0
Acker, W. L., Ill, Basic Procedures for Soil Sampling and Core
Drilling, Acker Drill Co., 246 p., 1974.
Bear, J., Dynamics of Fluids in Porous Media, American Elsevier,
764 p., 1972.
Bouwer, H., Groundwater Hydrology, McGraw Hill, 480 p., 1978.
Bouwer, H., and R. C. Rice, "A slug test for determining hydraulic
conductivity of unconfined aquifers with completely or
partially penetrating wells", Water Resources Research,
12, p. 423-428, 1976.
Bredehoeft, J. D., and S. S. Papadopulos, "A method for deter-
mining the hydraulic properties of tight formations",
Water Resources Research, 16, p.233-233, 1980.
Cooper, H. H., J. D. Bredehoeft, and I. S. Papadopulos,
"Response of a finite diameter well to an instantaneous
charge of water", Water Resources Research, 3, p. 263-269,
1967.
Freeze, R.A., and J.A. Cherry, Groundwater, Prentice Hall,
604 p., 1979.
Johnson, E. E., Inc., Groundwater and Wells, Johnson Division,
UOP, 440 P., 1975.
Lohman, S.W., Groundwater Hydraulics, U.S. Geological Survey
Professional Paper 708,70 p., 1972.
Neuzil, C.E., "On conducting the modified 'slug' test in tight
formations", Water Resources Research, vol. 18, no.2,
pp. 439-441, 1982.
4-1
-------
Papadopulos, S. S., J. D. Bredehoeft, and H. H. Cooper, Jr., "On
the analysis of 'slug test' data", Water Resources Research,
9, p. 1087-1089, 1973.
Sowers, G.F., "Rock permeability or hydraulic conductivity - an
overview", ir\ Permeability and Groundwater Transport, ed.
T.F. Zinunie and Co., O. Riggs, ASTM Special Technical
Publication 746, 1981.
Stallman, R. W., Aquifer-Test Design, Observation and Data
Analysis, TWRI, Chap. Bl, Book 3, U.S. Geological Survey,
U.S. Govt. Printing Office, Washington, D.C., 1971.
U.S. EPA, Test Methods for Evaluating Solid Waste; Physical/
Chemical Methods? Test Method 9100: Methods for Determining
Saturated Hydraulic Conductivity and Saturated Leachate
Conductivity; OSW, SW-846.
Walton, W. C., Groundwater Resource Evaluation, McGraw Hill,
664 p., 1970.
4.2 References for Section 2.0
Barari, A. and L.S. Hedges, "Movement of water glacial till",
Proceedings of the 17th International Congress of the
International Association of Hydrogeologists, pp.129-134,1985,
Bear, Jacob; Hydraulics of Groundwater, McGraw-Hill, Inc.., 1979
Bear, Jacob; Dynamics of Fluids in Porous Media, American Elsevier
Publishing Company, Inc., 1972
Bradbury, K.R., D.S. desaulniers, D.E. Connell, and R.G. Hennings,
"Groundwater movement through clayey till: Northwestern
Wisconsin, U.S.A.", Proceedings of the 17th International
Congress of the International Association of Hydrogeologists,
pp. 405-416, 1985.
Freeze, R. Alen and Cherry, John A., Groundwater, Prentice-Hall,
Inc., 1979
McWhorter, David B. and Sunada, Daniel K., Ground-Water Hydrology
and Hydraulics, Water Resource Publications, 1977
4-2
-------
Q5WEH POLICY DlRECTWt
9472 • 00
4.3 References for Section 3.0
GCA Technology Division, Inc., Technical Support for Development
and Analysis of Hazardous Waste Disposal Regulations, Draft
Background Document, March 1986.
U.S.EPA, RCRA Ground-Water Technical Enforcement Guidance Document
(Draft), 1985. Contact OWPE of the Office of Solid Waste and
Emergency for information on final publication date and availability
U.S.EPA, The Hydrologic Evaluation of Landfill Peformance (HELP)
Model, Volume 1.User's Guide for Version 1,Technical Resource
Document for Public Comment.
4-3
------- |