United States
                Environmental Protection
                Agency
 Risk Reduction  Engineering
 Laboratory
 Cincinnati, OH 45268
                Research and Development
 EPA/600/S2-91/022  July 1991
EPA       Project  Summary
                State-of-the-Art Field  Hydraulic
                Conductivity Testing  of
                Compacted  Soils
                Joseph O. Sai and David C. Anderson
                 The congressionally mandated per-
                formance standard for soil liners of haz-
                ardous waste management facilities is
                a hydraulic conductivity of 1 x 10* m/s
                or less. In response to this statutory
                requirement,  the U.S. Environmental
                Protection Agency  (EPA)  has issued
                guidance requiring that facilities dem-
                onstrate this hydraulic conductivity in
                field tests.
                 Hydraulic conductivity test methods
                currently used on soil liners were evalu-
                ated for their ability to meet the mini-
                mum requirements for field tests, i.e.,
                that the test be capable of measuring
                hydraulic conductivities of 1 x 1Q-9 m/s
                or less and that the values obtained be
                representative of the overall soil liner.
                Few methods are capable of meeting
                these minimum requirements, and even
                fewer are  both practical to use and
                rarely give false low values. Based on
                the  advantages of all the  methods
                evaluated, the best and most practical
                currently available technologies for
                evaluating hydraulic conductivity are
                large single-ring infiltrometers  and
                sealed double-ring infiltrometers. If cor-
                rection factors are needed to bring the
                values obtained with single-ring devices
                to below 1 x 10-8 m/s, confirmatory tests
                should  be conducted with sealed
                double-ring infiltrometers.
                 the size of  infiltrometers  used  on
                soil liners  should be at least 2 m2. In
                addition, at least three separate tests
                should be  conducted on each test fill
                to allow characterization of the spatial
                variability in the soil liner.
  A long-term study is needed to allow
 a comparative evaluation of candidate
 hydraulic conductivity testing devices.
 A large collection lysimeter should be
 incorporated into the study to  give the
 true overall hydraulic conductivity value
 with which other values should be com-
 pared.
  This Project Summary was developed
 by  EPA's Risk Reduction Engineering
 Laboratory, Cincinnati, OH, to announce
 key findings  of the research project
 that is fully documented In a separate
 report of the same title  (see Project
 Report ordering information at back).

 Introduction
  In the Hazardous and  Solid Waste
 Amendments (HSWA) of 1984, the Con-
 gress of the United States mandated that
 where compacted soil liners are used, their
 hydraulic conductivity shall be 1 x 10'9 m/s
 or less. In response to this statutory per-
 formance standard, EPA issued guidance
 requiring all proposed soil liners in haz-
 ardous waste  management facilities to
 demonstrate this hydraulic conductivity in
 field tests. The intent of this requirement
 was to obtain accurate and realistic evalu-
 ations of how  the compacted soil liner
 would perform under field conditions. Nu-
 merous studies have demonstrated that
 values  obtained in  laboratory hydraulic
 conductivity tests are not reliable indica-
tors of the performance of soil- liners un-
der  field conditions.  No single study has
been conducted,  however,  to document
the  range of field hydraulic conductivity
test methods capable of adequately evalu-
                                                                 Printed on Recycled Paper

-------
ating the field performance of a compacted
soil liner.
  Field  hydraulic conductivity tests  can
damage soil liners in several ways. Dam-
age can occur in the form of holes drilled
or trenches cut in the liner to facilitate the
installation  of  testing  equipment.  Also,
because field hydraulic conductivity tests
can take as long as sever.al weeks to
complete, both climatic events and weath-
ering processes can substantially damage
a soil liner.  Consequently, the EPA has
recommended that field hydraulic conduc-
tivity tests  be conducted on a test fill.
Both the test fill and the hydraulic conduc-
tivity tests  used  on the test fill should
provide a sound basis for meeting the
following objectives:
  1) Accurately measuring  hydraulic  con-
     ductivities of 1 x 10"9 m/s and lower.
  2) Obtaining hydraulic conductivity val-
     ues that  accurately   represent
     the properties of the full-scale soil
     liner.
  The hydraulic conductivity values ob-
tained must be representative of the full-
scale soil liner.   The representative el-
ementary volume (REV) of a soil liner is
the smallest volume above which the  vari-
ance no longer decreases significantly.  A
REV will be different for every liner and be
highly dependent on the natural variability
of the soil material and the  level of quality
assurance  exercised  during  construction
of the soil  liner.  At least three field hy-
draulic conductivity tests will  be required
to define the REV for a given soil liner.
  The available  field hydraulic conductiv-
ity  test methods for soil liners discussed
herein are currently used and  readily avail-
able for determining the hydraulic conduc-
tivity of soils compacted in the field.

Field Hydraulic Conductivity
Methods
   A wide variety of field hydraulic conduc-
tivity methods have been published in the
scientific and engineering literature. Many
of  these methods  are  not adequate for
evaluating  soil liners because they can
neither measure very low  hydraulic  con-
ductivities  nor evaluate a large enough
area to give values representative of the
overall liner. Other methods are adequate
to obtain representative values on low hy-
draulic conductivity soils, but they are  com-
plex and time consuming or require a  large
amount of  time from skilled equipment
operators.

Air-Entry Permeameters
   Air-entry permeameters give rapid field
measurements of vertical  hydraulic  con-
ductivity in initially unsaturated soil.  This
method uses Darcy's fundamental law for
water flow through soil to determine soil
hydraulic conductivity above a water table.
  The air-entry permeameters method can
rapidly measure field hydraulic conductiv-
ity values as low as 1 x 10'9  m/s.  One
hour or less is needed to run the test,
depending on the temperature of the wa-
ter and the hydraulic conductivity of the
soil.  A major shortcoming of this method,
however,  is that it tests only small areas
of soil (0.3 m2) and therefore may not
reflect the effect of  soil macropores on
hydraulic  conductivity.   Soil rnacropores
that are widely  spaced in a soil liner can
have pronounced effect on hydraulic con-
ductivity.
  Air-entry permeameters can  be used to
obtain the vertical hydraulic conductivity
of one lift in a compacted soil  liner or the
entire thickness of  soil liner.  The data
may be  inaccurate,  however, either be-
cause of inadequate sealing between the
cylinder and the soil  or because of errors
in determining the exact locations of the
wetting front. Air-entry permeameters are
easy  to transport and  to use.  Also, the
method is not  labor-intensive (two  men
can set up the equipment and make the
measurements).

Borehole Methods
  Several borehole  methods  have been
used  to measure the hydraulic conductiv-
ity of soils.  Borehole  permeameters are
essentially in-hole constant-head or fall-
ing-head  permeameters.   The  methods
involve measuring the steady-state  infil-
tration rate of water into unsaturated soil
from a cylindrical borehole. Two borehole
methods that have been used to test the
hydraulic conductivity  of soil liners are
Guelph and Boutwell permeameters. The
Guelph permeameter uses a Mariotte si-
phon to maintain a constant head of water
within  the  borehole.   The Boutwell
permeameter involves an evaluation of the
flow into a cased borehole and an evalua-
tion of subsequent flow after an uncased
extension has been added to the hole.
  Guelph permeameters have not been
widely used to determine hydraulic  con-
ductivity  on  compacted  soil liners.
Studies conducted on  low hydraulic  con-
ductivity soils indicated, however, that the
instrument is capable of measuring values
as  low as 2 x  10'9 m/s.   The Guelph
permeameter only evaluates a small  area
of soil (3 x 10-1 to 2 x 10'3 m2).   For this
reason, the method is unlikely to give val-
ues that  are representative of the overall
field hydraulic conductivity.
   Boutwell permeameters can  measure
hydraulic conductivities of 1  x  10'9 m/s
and less.  The volume of soil tested in a
single  borehole,  however,  is  relatively
small.  Because of the small size of the
test area, soil macropores and other flaws
in  soil  liner construction may be missed
with this method.  As  a result, hydraulic
conductivity determined by  this method
may not be representative of  the actual
field value.
   Multiple  tests could be  conducted to
evaluate a larger aggregate'area. For sev-
eral reasons, however, running many small
tests may  not  achieve the desired test
objective  of obtaining a  representative
measure of the field hydraulic conductiv-
ity. For example, even if many tests were
conducted, the small scale of each indi-
vidual test would greatly increase the prob-
ability  that through-going  macropores
would  be truncated.  Such truncation of
through-going macropores  would signifi-
cantly  reduce the  hydraulic conductivity
below the actual field value. Other prob-
lems include the large amount of potential
smearing per unit volume pf soil and the
difficulty that may be encountered in dis-
tinguishing between defective test results
and results  reflecting the value  of a small
area of liner with  a macropore.

Porous Probe Permeameters
   Porous probe  permeameters typically
consist of  a cone-shaped  porous  probe
that is  either pushed  or driven into the
soil.   One  such commercially available
porous probe, the BAT permeameter,' can
measure  hydraulic conductivities of  1 x
10"9 m/s or less.   Because of the small
volume of  soil  tested :by   the  BAT
permeameter, the  method is  unlikely to
yield values consistently representative of
the overall  hydraulic conductivity of the
soil being tested.

Ring Infiltrometers
   In ASTM Method D 3385, double-ring
cylinders are used to  determine the rate
of infiltration of  water into soils.   This
method has been  widely used for  more
than a decade for the evaluation of both
percolation  rate in septic fields and  infil-
tration rate in soils to be irrigated.  Two
open cylinders, one inside the  other, are
driven into the soil and partially filled with
water.  A constant water  level is main-
tained on the soil by continuously adding
water.  The volume of water added  in a
given period is used to determine the rate
of infiltration.
 • Mention of trade names or commercial products does
  not constitute endorsement or recommendation for
  use.

-------
   The  ASTM  double-ring infiltrometers
 should  not be used to measure infiltration
 rates on compacted soil liners.   This
 method is both difficult to use and  unreli-
 able in  soils with a hydraulic conductivity
 less than about 1 x 10'8 m/s. Also, loss of
 water due to evaporation  may  be  higher
 than the quantity of water  permeating the
 soil.
   Modifications  have  been made to the
 basic ASTM double-ring  infiltrometer  to
 make it more sensitive for measuring low
 infiltration rates.  These modifications in-
 clude the use of a  large  outer ring and
 sensitive devices for measuring  water
 level. With these modifications, the ASTM
 double-ring device has been used to mea-
 sure infiltration rates  of 9.7(± 8.1) x 10'10
 m/s. The ability of the test to obtain  repre-
 sentative values is questionable, however,
 because of the relatively small size of the
 inner ring (0.3 m in diameter). Macropores
 that can affect the infiltration rate by or-
 ders of magnitude could  have a  wider
 average spacing than  the diameter of the
 inner ring.
   Box infiltrometers use a sealed box and
 a relatively complex  standpipe arrange-
 ment to measure infiltration into soil in a
 0.36-m2 area.  They  have  been used to
 measure  hydraulic conductivities as low
 as 5 x 10'10 m/s.
   A single-box infiltrometer coverage (0.36
 m2) is not sufficient to determine the over-
 all field  hydraulic conductivity. One study
 of box infiltrometers used arrays of four to
 allow definition of both  the overall field
 hydraulic conductivity and the spatial vari-
 ability of the liner.  The aggregate area
 evaluated by the infiltrometers  (1.44 m2)
 appeared to be sufficient to obtain repre-
 sentative  hydraulic conductivity values.
 The test site was small (8 m2),  however,
 and it required greater effort per unit area
 to produce  a uniform soil liner than  is
 typical for full-scale liners.  Consequently,
 although an  aggregate area of 1.44 m2
 may have been  sufficient to characterize
 the spatial variability  of  the test site,  a
 larger area  probably  would have to be
 tested  under conditions  more  typical  of
 full-scale liners.  The  box  infiltrometer  is
 relatively difficult to use  and would  defi-
 nitely require skilled personnel to install,
 monitor, and collect and analyze data.
   A single-ring infiltrometer consists of  a
 metal cylinder (20 to 60 cm in diameter),
 which is pressed or driven into the soil
 (Figure  1).   Infiltration is  measured by
 ponding water inside the cylinder and then
 (1) measuring the rate at which the free
surface  falls,  or (2) measuring the rate at
which water must be added to maintain  a
constant depth in the cylinder.
   Large single-ring infiltrometers can mea-
 sure hydraulic conductivity values as low
 as 1 x 10'9 m/s.  If the ring is of sufficient
 diameter,  the  hydraulic conductivity val-
 ues obtained should be representative of
 the overall soil  liner.   Until there is a
 definitive data base documenting the mini-
 mum ring diameter needed to obtain these
 representative  values, a total area  of  at
 least 2 m2 within the ring is recommended.
   Errors can be introduced into the hy-
 draulic conductivity values as a result  of
 lateral flow and evaporative losses. There
 is no danger of falsely concluding that a
 soil liner meets  a specific  performance
 standard,  however, as long  as  the
 uncorrected hydraulic conductivity values
 are less than 1x1O'9 m/s. If a correction
 factor must be used to  reduce hydraulic
 conductivity below  1 x 10/9 m/s, additional
 tests should be conducted with  either a
 double-ring infiltrometer or by using mea-
 sures to eliminate evaporative losses.
   Single-ring infiltrometers have the ad-
 vantage  of being simple  and  inexpensive.
 Care should be taken, however, to make
 sure that thermal expansion  of the water
 does not yield false low values.  Other
 possible sources of false low values are
 the interception of rainfall by the open ring
 or  incorrect calculation  of the  hydraulic
 gradient.  These  sources of  error can  be
 avoided by using a sealed ring and a free-
 draining  layer  beneath the known  thick-
 ness of a soil liner, respectively.
   Sealed double-ring infiltrometers (SDRI)
 use a sealed ring  or box that measures
 infiltration into a relatively large area.  The
 SDRI has a sealed inner ring that permits
 measurement of low infiltration rates and
 minimizes problems with temperature fluc-
 tuations  and evaporation.  It  is relatively
 easy to operate, but as with all infiltration
 ring devices, it  must be carefully installed
 to ensure that no leakage occurs around
 the edges.
  The SDRI consists of an inner ring and
 an  outer ring (Figure 2).  The fiberglass
 inner ring has a sloped top that extends
 only 12 cm above  the liner surface, and
the outer ring is used to pond water around
the inner ring to ensure only  vertical per-
colation is  measured.  Flow is measured
during an infiltration test  by using a flow-
 measurement bag attached  to  the
infiltrometer.  Any  water flow out of the
infiltrometer into the ground is replaced by
water from the bag. Flow measurement is
initiated by filling the bag  with a known
weight of  water,  connecting it to the
infiltrometer, and periodically retrieving and
reweighing it.
  The SDRI can measure hydraulic con-
ductivities of less than 1  x 10'9 m/s.  The
 SDRI also is available commercially in  a
 relatively large size (the inner ring covers
 an area of 2.3 m2).  Tests conducted with
 SDRI's appear to confirm that the equip-
 ment can yield  values  representative of
 the overall soil liner.
   Although SDRI tests require  more in-
 stallation time than many methods, peri-
 odic recording of time and measuring bag
 weight are the only tasks performed after
 the equipment is  installed.   The SDRI
 appears to yield high quality data with few
 possibilities for yielding false low values.

 Collection Lysimeters
   Collection lysimeters are placed beneath
 the compacted soil liner to collect liquid
 percolating through the liner.  A pipe is
 often installed at the low end of the lysim-
 eter to collect and move any accumulated
 liquid to an access point where the liquid
 can be measured.  Lysimeters have been
 used to monitor the quantity and quality of
 the leachate from landfills in Wisconsin
 and Canada.
   Collection lysimeters can measure hy-
 draulic conductivities of less than 1 x 10'9
 m/s. If large enough, lysimeters  can also
 yield hydraulic conductivity values that are
 representative of the overall soil liner. One
 study showed a  19.4-m2 collection lysim-
 eter to be large enough to yield represen-
 tative values.  The main drawback of this
 method is that it can take months to ob-
 tain steady-state hydraulic conductivity val-
 ues.  Unlike infiltrometers,  which give ini-
 tially high  values that decrease to a low
 steady-state value, collection lysimeter val-
 ues typically begin very low and gradually
 build up to a higher steady-state  value.
 Consequently, the testing period may well
 interfere with facility construction and the
 adequacy of a particular soil liner design
 may not be known until the end of a long
test period.
   Tests  comparing  collection  lysimeters
 and large  infiltrometers have  shown that
values are relatively close.  Therefore, if a
collection lysimeter is desired, concurrent
testing  of the  soil  liner  with  large
infiltrometers is suggested.  This proce-
dure  could save  months  of  waiting for
results definitive enough to begin construc-
tion of a facility.
   Constructing a collection  lysimeter is
time-consuming and requires skilled per-
sonnel.  Special  care must be taken to
avoid damaging the material from which it
is  constructed  underlying  the collection
field.  Because such damage could result
in  false low hydraulic conductivity values,
concurrent testing of the soil  liner with
large infiltrometers is  recommended. Ulti-
mately,  a  large collection  lysimeter has

-------
                                                      V     V
Figure 1. Open Single-Ring Infiltrometer.
            -OUTER RING WALL
                                                                            
-------
the potential for yielding the most accu-
rate values for quantification  of the  vol-
ume and rate of liquid that moves through
a compacted soil liner.  In practice, how-
ever, the response time of a  lysimeter
limits its utility for short-term tests.

Other Apparatus
  The  applicability of  both  velocity
permeameters   and   porous   plate
infiltrometers for measuring the  hydraulic
conductivity of compacted soils is being
studied.  Current data are not sufficient to
determine the adequacy of either method,
but both methods show promise.
  Various methods are also available for
obtaining the hydraulic conductivity of soil
cores in the laboratory.  Laboratory meth-
ods generally do  not  provide  a reliable
indicator of  field performance  of clay lin-
ers.
  Laboratory hydraulic conductivity tests
can measure very low hydraulic conduc-
tivity values; however, using these values
to represent field conditions presents sev-
eral  problems.   Because  laboratory
samples used to determine hydraulic con-
ductivities are typically too small to have a
representative  distribution  of   the
macropores present in the field, the  val-
ues derived are one to  three orders of
magnitude lower than the actual field  val-
ues.

Conclusions
  Widely  varying methods  are  currently
being used to evaluate the field  hydraulic
conductivity of soil liners.  Only a few of
these methods, however, can reliably meet
the following requirements for evaluating
soil liners:
  1) A method should be capable of ac-
     curately measuring hydraulic conduc-
     tivities of 1 x 10'9 m/s and lower.
 2) The  values obtained should be rep-
     resentative of the overall  hydraulic
     conductivity of a soil liner.
  Of the few methods capable of meeting
these requirements, even fewer are rela-
tively simple to use, rarely give false  low
values, and  provide definitive results in  a
reasonable time frame.  Key findings of
each  method evaluated are summarized
here.
  Air-entry  permeameters,   velocity
permeameters, porous plate infiltrometers,
borehole  methods,  and  porous probe
methods,  can give rapid  field  measure-
ments of  vertical  hydraulic conductivity,
but they may not test a large enough area
to give values representative of the over-
all  soil  liner. The ASTM  double-ring
infiltrometers use an outer ring to ensure
that the inner ring measures only the ver-
tical hydraulic conductivity of a soil.  This
method cannot measure hydraulic  con-
ductivity values of less than  1 x  1 Cr8 m/s
or obtain values representative of  an over-
all soil liner.   The primary disadvantages
of box infiltrometers are that they  are diffi-
cult to use and that skilled personnel would
definitely be  required for installing, moni-
toring, and collecting and analyzing data.
  Single-ring  infiltrometers should cover
an area of at least 2 m2 and care should
be taken not to rely on correction factors
to reduce the  hydraulic conductivity val-
ues below 1  x 10'9 m/s.  Sealed  double-
ring infiltrometers have received substan-
tial field testing and have a demonstrated
capability of measuring hydraulic  conduc-
tivity  values  below 1  x ICr9  m/s  and  of
obtaining values representative of the over-
all soil liner.   The method requires more
installation time than many methods, but
it has the advantages of few ambiguities
in the experimental method and few  pos-
sibilities for yielding false low values.  Col-
lection lysimeters can  be time-consuming
and may require  skilled personnel.  Ulti-
mately, this method  has the  potential for
yielding the most accurate values for hy-
draulic conductivity of a soil liner.

Recommendations
  To  maximize the probability of obtaining
a hydraulic conductivity value that is rep-
resentative of the overall soil liner, a test
method should have  the following quali-
ties:
  1)  The ability to measure hydraulic con-
      ductivities of less than 1 x 10'9 m/s.
  2)  Minimal requirements for skilled per-
      sonnel during installation, operation,
      data acquisition, data reduction, and
      interpretation of results.
  3)  Few ambiguities in the experimental
      method and few possibilities for yield-
      ing false low values.
  4)  Inexpensive enough to allow at least
     three replicate tests to determine
     spatial variability  of field  hydraulic
     conductivity.
  5) Sufficient area of coverage for each
      replicate test to ensure that a statis-
     tically  sound  average  number  of
     macropores per unit area is covered.
  6) Sufficiently short  time  required  to
     conduct each test to ensure that it is
     practical for use.
   Not all of these qualities are quantifi-
 able, given the current state of knowledge
 on hydraulic conductivity test methods and
 soil liners.   It  is possible,  however, to
 pinpoint where additional study is needed
 and to suggest the best currently avail-
 able methods.
   For a better definition of the test meth-
 ods that would maximize the probability of
 obtaining representative hydraulic conduc-
 tivity  values for soil liners, a  long-term
 soil-liner study is recommended that would
 incorporate a large collection lysimeter (to
 give true overall hydraulic conductivity) and
 arrays of candidate hydraulic conductivity
 testing devices.  The results could be used
 to define the reliability  of the testing de-
 vices and to characterize their sensitivity
 to spatial variability within the  soil  liner.
 Ideally, several sized versions of each test
 method should  be used in each array to
 obtain a better understanding of the mini-
 mum size requirement for the  test de-
 vices.
   Based  on the advantages and disad-
 vantages of the methods reviewed in this
 document, the following  are recommended
 as the  best  and most  practical currently
 available technologies for evaluating field
 hydraulic conductivity:
  1) A single-ring infiltrometer covering
     an area greater than 2 m2.
  2) A sealed  double-ring  infiltrometer
     with an inner ring covering an area
     greater than 2 m2.
   Until conclusive studies are completed,
 correction  factors (such as those com-
 monly  used with single-ring  devices)
 should not be relied on  to reduce hydrau-
 lic conductivity values below 1 x 1Cr9 m/s.
 Also,  correction  factors for evaporative
 losses  should  be  avoided.   If  the
 uncorrected hydraulic conductivity values
 are higher that the  maximum allowable
 value (1  x  10'9 m/s), confirmatory tests
 should be conducted  with sealed double-
 ring infiltrometers.
   Spatial variability of the soil  liner must
 be characterized to  ensure that the hy-
 draulic conductivity values  obtained  are
 representative.   At least three field  hy-
 draulic conductivity tests should  be  con-
 ducted, and the values obtained from each
of the three should be 1  x 10-9 m/s or less.
  The  full report was submitted  in fulfill-
 ment of Contract No. 68-03-3413 by  PEI
Associates, Inc., under the sponsorship of
the U.S. Environmental Protection Agency.
                                                                               •&U.S. GOVERNMENT PRINTING OFFICE: 1991 - 548-028/40049

-------

-------

-------
   Joseph O. Saiand David C. Anderson are with K.W. Brown & Associates, Inc.,
     College Station, TX 77840.
   Walter E Grube, Jr.,  is the EPA Project Officer (see below).
   ThQ complete report, entitled "State-of-the-Art Field Hydraulic Conductivity Testing of
     Compacted Soils, "(Order No. PB91-206243/AS; Cost: $17.00, subject to change)
     will be available only from:
           National Technical Information Service
           5285 Port Royal Road
           Springfield, VA 22161
           Telephone: 703-487-4650
   The EPA Project Officer can be contacted at:
           Risk Reduction Engineering Laboratory
           U.S. Environmental Protection Agency
           Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati, OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA PERMIT NO. G-35
  Official Business
  Penalty for Private Use $300
  EPA/600/S2-91/022

-------