Geosynthetic Research Institute

475 Kedron Avenue GRI

Folsom, PA 19033-1208 USA GEI GII
TEL (610) 522-8440 GSI
Drexel
U N I V E R S I T Y
FAX (610) 522-8441
GAI GCI

A GRI White Paper

on

The Questionable Strategy of Soil-Only

Landfill Covers

by

Robert M. Koerner, Ph.D., P.E.

Director and Professor

Geosynthetic Research Institute

475 Kedron Avenue
Folsom, PA 19033-1208

GRI Report #28

August 7, 2002
 Geosynthetic Research Institute
White Paper

on

“The Questionable Strategy of Soil-Only Landfill Covers”

Synopsis

A large scale field study of percolation through six different landfill final cover

cross sections at Sandia National Laboratories in Albuquerque, New Mexico has

prompted this white paper. The primary focus of the Sandia field study was to evaluate

three soil-only strategies; namely, capillary, anisotropic and thick soil cross sections. As

a secondary focus they were also compared to three traditional cross sections; two

contained geomembranes and the third consisted of a low-permeable compacted soil.

However, the two sections which included geomembranes had holes deliberately

punched in the geomembranes (and perhaps the geosynthetic clay liner as well) during

installation. In contrast, none of the other materials at any of the test plots were

comparably damaged during construction. The negative implications of this “flawed”

field study as to the performance of geosynthetics in construction of final covers for

landfills are discussed in light of the data from the Sandia Test Plots.

Background

The final cover of a closed landfill containing either hazardous or nonhazardous

waste is a challenge to a designer due to a number of inherent variables; for example,

(i) type and extent of the waste mass,

(ii) hydrologic conditions at the site,
(iii) condition and/or existence of the base liner system,
(iv) sensitivity of the surrounding environment,

-1-
 (v) connectedness of the landfill to ecological pathways,
(vi) design lifetime of the proposed final cover, and
(vii) potential future beneficial use of the site.

As a result, landfill final covers are always site-specific designs. Yet, there are some

generalities related to the design of the specific layers that a designer must consider.

Figure 1(a) presents the identification of the general layers, and Figure 1(b) presents

the geosynthetic materials that are often used to replace or augment some, or all, of the

natural soil materials that might be used. To be sure, geosynthetics can and have

played a critical and very positive role in providing environmentally safe and secure final

covers at hundreds (perhaps even thousands) of closed landfill facilities. See, for

example, the data base accumulated by Othman, Bonaparte and Gross (1997).

(a) Essential layers of a final cover system

(b) Geosynthetic alternatives for natural soils

Figure 1 – Cross Sections of Final Covers for Landfills, after Koerner and Daniel, 1997

-2-
It is readily acknowledged, however, that such final covers are not inexpensive as the

following generalized cost estimate indicates.

Table 1 – Estimated Costs of Engineered Landfill Final Covers, Koerner (2001)

Item Description Cost/Acre Cost/Hectare

exploration soundings/test pits 15,000 31,000
design plans/specifications/permits 25,000 62,000
construction earthwork/geosynthetics 70,000 172,000
inspection MQA/CQA 10,000 25,000
guarantees insurance/bonding 20,000 50,000
maintenance vegetation/fencing/signage 10,000 25,000
TOTAL $150,000 $365,000
Note: These are approximate costs; they are extremely site specific and can vary by as
much as 50% from site-to-site.

In assessing the specific items in the above table it is seen that the physical

construction item represents approximately half of the total cost. As such, construction

(including materials) is logically a major target insofar as possible cost reduction is

concerned. Depending on the number and types of geosynthetics (recall Figure 1b) and

the size and location of the project, one could target the geosynthetics as being a

candidate for removal from the cross section (environmental safety and security issues

aside) so as to decrease the overall cost of the final cover. It appears as though this is

the approach that is being taken under the title of “alternative landfill covers”, aka

inexpensive soil liners without geosynthetics. There appears to be two different

approaches to soil-only covers; capillary (or anisotropic) barriers and thick monolayer

barriers. Each will be briefly described.

Capillary (or Anisotropic) Barriers

A fairly recent development is the use of a layer of fine-grained soil overlying a

layer of coarse-grained soil to form a capillary barrier. Typically, this is sand overlying

gravel where the differences in hydraulic conductivity (or permeability) are 3 or 4 orders

-3-
of magnitude. A slight variation is called an anisotropic barrier. The idea of both

capillary and anisotropic barriers is as follows. Soil moisture in the subsurface reaches

equilibrium when the soil water potential is the same throughout. If a layer of fine- and

coarse-grain soil are in equilibrium and there is no movement of water between the

layers, the two layers will have the same soil water potential. For a given soil water

potential, a coarse-grained soil will tend to have a much lower water content, i.e., be

much drier, than the overlying fine-grained soil. Furthermore, the permeability of

unsaturated soil decreases exponentially with decreasing water content because flow

paths through thin films of water coating the soil particles in a relatively dry soil are

extremely tortuous. A dry gravel is actually much less permeable to small quantities of

water than a moist sand.

Thus, if the subsoil remains unsaturated, a fine-grained soil overlying a coarse-

grained soil will tend to function with the uppermost soil layer retaining nearly all of the

soil moisture and the underlying layer serving as a de facto barrier to water percolation

due to its dryness. These two distinct soil layers are called a capillary barrier system.

Note, however, that in a capillary barrier, lateral movement of water in the fine-grained

soil occurs in the unsaturated state. For this reason the upper soil layer is sometimes

referred to as a wicking layer.

There are a number of concerns with capillary barriers. One is that the fine-

grained soil must not be allowed to migrate downward over time into the underlying

coarse-grained soil. Obviously, this would completely destroy the concept and system’s

functionability. A geotextile, used as a separator, should be considered for placement

beneath the fine-grained soil and above the coarse-grained soil. For extremely long

-4-
service lifetimes, the durability of the geotextile must be assessed for this application.

Alternatively, a graded granular soil filter might be used instead of a geotextile but this

tends to mitigate the differences in permeability upon which the concept is based. A

second concern is over sloping portions of the system where the wicked water in the

fine-grained soil can accumulate and eventually breakthrough into the coarse-grained

soil. The third concern is over periods of high (relatively speaking) concentrated

precipitation or snow-melt. In such cases, the capillary barrier concept may cease to

function, as least temporarily, as the coarse-grained soil becomes moist and loses it

water impeding capability. Clearly, these concerns are in need of investigation using

large scale or full scale test plots as is being done in the study to be described.

Irrespective of the concerns that were just mentioned, it is generally agreed that

capillary barriers can only be considered for arid or semi-arid areas where precipitation

is low and does not occur in short term increments, i.e., the precipitation is relatively

evenly spaced throughout the year. Also, note that the terms arid and semiarid are

quantified indexes. Thornthwaite (1948) uses a moisture index defined as annual

precipitation minus evapotranspiration in units of inches per year. In doing so, an arid

climate is between -60 and -40. A semiarid climate is between -40 and -20. This

definition restricts the applicability of capillary barriers in the United States to the

southwest for arid areas, and other western areas (except for the Pacific coast) for

semi-arid areas.

Thick Monolayers Barriers

Monolayer barriers are covers that include a thick layer of fine-grained soil

generally covered with a layer of vegetated topsoil. They are also called

-5-
evapotranspiration covers. This type of thick cover encourages water storage and

enhances evapotranspiration year-round, rather than just during the growing seasons.

The soil allows water storage, which, when combined with the vegetation, will increase

evapotranspiration. Monolayer barriers exploit two characteristics of fine-grained soils;

(i) their large soil moisture storage capacity when unsaturated, and (ii) their low

saturated permeability relative to coarse-grained soils (Morrison-Knudsen, 1993). The

soil’s low saturated permeability limits infiltration through the surface during rainfall or

snowmelt. The soil’s high moisture storage capacity makes it capable of storing water

that does infiltrate until it can later be removed by evapotranspiration. The barrier must

be sufficiently thick, however, such that changes in water content do not occur near its

base, i.e., all changes in soil moisture storage must occur in the upper portion of the soil

barrier. Otherwise, water will percolate into the underlying waste. The necessary

thickness is a function of the amount of precipitation received, the unsaturated hydraulic

properties of the soil, and the rate at which water can be removed by

evapotranspiration. Monolayer barriers are constructed from silty sands, silts, and

clayey silts. These soil barriers can be cost effective when large quantities of fine-

grained soil requiring little processing is available on site.

Geologic Associates (1993) describes a field study conducted to assess the

performance of a thick soil barrier used as final cover for a landfill in Southern

California. The barrier was 2-m (6.6-ft)-thick and was constructed from a clayey silt.

Water movement was limited to the upper 0.6 m of soil; no changes in water content

were observed at the base of the barrier. The data indicated that the water content of

the upper soil layers increased rapidly after rainfall and then decreased as water was

-6-
removed by evapotranspiration. However, the data collection time was only a few years

and subsequent data is not available to our knowledge.

Concerns over thick monolayer soil barriers focus on preferential flow pathways

which can develop in monolayer barriers as a result of (i) desiccation cracking, (ii) root

growth and penetration, and (iii) burrowing animals. Sufficient data about the

performance of monolayer barriers have not been gathered so far from which to judge

their reliability and effectiveness in this regard, Benson and Khire (1995). They feel that

field tests including large-scale measurements of percolation are needed before

definitive conclusions regarding monolayer barriers can be drawn.

The Sandia Test Plots

In an attempt to evaluate both alternative and traditional (geosynthetic-related)

final landfill covers, the U. S. Department of Energy funded a field study at Sandia

National Laboratories (Kirkland Air Force Base) in Albuquerque, New Mexico, see

Dwyer (1998 and 2001). While clearly stated in the available reports that the study “…

is not intended to showcase any particular cover system”, the bias included in the field

deployment of the geosynthetics did precisely that. The results in the traditional

(geosynthetic-related) test plots did not perform as well as would be expected and in

one case did not perform as well as the soil-only test plots. The text to follow presents

the details of the Sandia Test Plots.

The test plots of Sandia are each 13 m wide by 100 m long. Half of each length

faces east (which includes sprinkler systems) and the other half faces west (which

includes passive monitoring). All slopes are at 5%. Table 2 presents the essential

details of the six test plots.

-7-
 Test Plots 1, 2 and 3 are considered to be traditional, in that the barrier layers are

450 mm of compacted clay, a composite liner consisting of a geomembrane over 600

mm of compacted clay, and another composite liner consisting of a geomembrane over

a geosynthetic clay liner, respectively. Since they follow U. S. EPA guidance, these test

plots are called Subtitle D (Soil), Subtitle C (GM/CCL) and Subtitle C (GM/GCL). Of

overriding importance is the fact that the geomembranes in Test Plots 2 and 3 were

purposely damaged by the incorporation of eight 1 cm2 defects (puncture holes)

punched through them during construction. If, and how far, the steel rods making the

punctures penetrated the underlying GCL and/or CCL is not known. Paradoxically, the

drainage system for data collection beneath all six of the test plots consists of geonet

drainage composites and geomembrane liners with no holes punched in them ! The

collection of potential leakage is apparently more important than the performance of the

geomembrane-related test plots. As far as the soil-only test plots, Test Plots 4 and 5

are both variations of a capillary barrier system (called capillary and anisotropic,

respectively) and Test Plot 6 is a thick soil monolayer.

Table 2 – Materials Used to Construct the Sandia Cover Test Plots

Number Characterization Layers (Top-to-Bottom)

1 Subtitle D Barrier (Soil) Topsoil (150 mm); Compacted Soil (450 mm)
2 Subtitle C Barrier Topsoil (600 mm); Geotextile; Sand (300 mm);
(with GM/CCL) Geomembrane (with Holes); Compacted Clay
(600 mm)
3 Subtitle C Barrier Topsoil (600 mm); Geotextile; Sand (300 mm);
(with GM/GCL) Geomembrane (With Holes); Geosynthetic Clay
Liner
4 Capillary Barrier Topsoil (300 mm); Sand (150 mm); Gravel (220
mm); Barrier Soil (450 mm); Sand (300 mm)
5 Anisotropic Barrier Topsoil (150 mm); Soil (600 mm); Fine Sand
(150 mm); Gravel (150 mm)
6 Thick Monolayer Barrier Topsoil (150 mm); Soil (900 mm)
(Evapotranspiration)

-8-
 As mentioned previously, each of the six test plots were underlain by a

underdrain system consisting of a geonet composite and a geomembrane. The

collected percolation water from the geonet composite was routed to a collection system

and measured accordingly.

To our knowledge there were no intentional defects placed in any of the soil

material test plots, i.e., in Test Plots 4, 5 or 6. For example, there were no soil

nonhomogenities, preferential flow paths, different compactive energies, different

placement moisture contents, etc., purposely induced into the soil materials. Thus, one

can expect at the outset that Test Plots 2 and 3 (with geomembrane holes) will be

overestimated insofar as their percolation (leakage) is concerned and thus behave

relatively poorly in light of what would be expected with proper construction. Also Test

Plot 1 with its relatively high permeability is expected to behave poorly. It might be noted

that many regulatory agencies and designers discount this particular cross section

completely. Thus, before one even looks at the test results, the outcome is essentially

known, i.e., that the alternative landfill covers will be favored with much lower leakage

than the Subtitle D (Soil) section and possibly the Subtitle C sections (with holes in the

geomembranes) as well.

The Sandia test results are presented in Table 3 for the 3-year time period from

May 1997 through June 2000. According to Dwyer (2001), the first year was apparently

quite wet, while the second two years were extremely dry.

-9-
Table 3 – Measured Percolation Rates in Units of mm/year Through the Sandia Landfill
Cover Test Plots, (from Dwyer, 2001)

Precipitation Plot #1 Plot #2 Plot #3 Plot #4 Plot #5 Plot #6

Volumes Subtitle D Subtitle C Subtitle C Capillary Anisotropic Thick
Year Collected (L) (Soil) (GM/CCL) (GM/GCL) Barrier Barrier Monolayer
1997 154,585 10.62 0.12 1.51 1.62 0.15 0.22
(May 1-Dec. 31)
1998 169,048 4.96 0.30 0.38 0.82 0.14 0.44
1999 130,400 3.12 0.04 4.31 0.85 0.28 0.01
2000 28,151 0.00 0.00 0.00 0.00 0.00 0.00
(Jan. 1-June 25)
Average 4.82 0.13 1.81 0.87 0.16 0.19

Commentary by Dwyer (2001) on these test results is as follows:

Test Plot 1 – “by far the worst design”

Test Plot 2 – no comment

Test Plot 3 – “appears to be degenerating with time”

Test Plot 4 – no comment

Test Plot 5 – no comment

Test Plot 6 – “appears to be leading the way”

Dwyer (2001) goes on to say about Test Plot 6, “This test reveals that in dry

environments a well-designed simple soil cover is not only the cheapest alternative but

also the most effective at controlling infiltration”. Thus, even before the anticipated 5-

year results are recorded the conclusion is finalized, as could have been anticipated on

the basis of the original experimental design and its flawed construction insofar as holes

in the geosynthetics are concerned.

Summary

The “flawed” large scale field study funded by the U. S. Department of Energy at

Sandia National Laboratories in Albuquerque, New Mexico has clearly driven the need

for this GRI white paper. Had the study focused on soil-only landfill covers and

- 10 -
investigated the nuances of capillary, anisotropic and thick monolayer barriers by

themselves it could have been a valuable contribution to the body of knowledge in this

particular application. Unfortunately, the comparison to traditional soil covers with

purposely flawed geomembranes (and perhaps GCLs and CCLs as well) serves

absolutely no purpose. Geomembranes need not have holes and with proper

construction quality control (CQC) superimposed with proper construction quality

assurance (CQA) holes have been often completely eliminated. Obviously, holes can

occur. However, if flawed construction practices are the target of the investigation, all

other materials should have deficiencies purposely installed in them as well. This

includes such well known soil inconsistencies as nonhomogeous materials, differences

in placement moisture contact and compactive energy, differences in varying soil layer

thicknesses and uniformity, etc. As far as Test Plot 1 with its relatively thin soil layer of

high permeability, it appears in the test program as a “red herring”. Most

knowledgeable people (regulators and designers) know this is an inadequate final cover

concept and simply ignore its existence.

In summary, it is strongly believed that if one desires an environmentally safe

and secure final cover for a landfill it must include a geomembrane or a geosynthetic

clay liner, and preferably both in the form of a composite barrier. As shown in the U. S.

EPA study, (Othman, Bonaparte and Gross, 1977) geosynthetics clearly work!

Thus, the results of the Sandia Test Plot study are of interest insofar as the soil-

only cross sections are concerned, but are completely irrelevant with respect to those

incorporating geosynthetics. In this regard, it is disappointing that a federal agency

should sponsor and conduct such a comparative study, particularly when it appears to

- 11 -
be contrary to the recommended U.S. EPA and most State EPA regulations for final

covers of closed landfill facilities.

References

Benson, C. H. and Khire, M. V. (1995), “Earthen Covers for Semiarid and Arid
Climates,” Landfill Closures—Environmental Protection and Land Recovery, ASCE,
Geotechnical Special Publication, No. 53, R. J. Dunn and U. P. Singh, Eds., New York,
NY, pp. 201-217.

Dwyer, S. F. (1998), “Alternative Landfill Covers Pass the Test,” Civil Engineering,
September, pp. 23-26.

Dwyer, S. F. (2001), “Finding a Better Cover,” Civil Engineering, January, pp. 59-63.

Geologic Associates (1993), “Evaluation of Unsaturated Fluid Flow, Coastal Sage Scrub
Habitat Area, Coyle Canyon Final Cover System, Orange County, CA,” Report prepared
by Geologic Associates for Orange County Integrated Waste Management Department.

Koerner, R. M. (2001), Presentation to Department of Waste Management,

Commonwealth of Virginia.

Koerner, R. M. and Daniel, D. E. (1997), Final Covers for Solid Waste Landfills and
Abandoned Dumps, ASCE Press, Reston, VA, 256 pgs.

Morrison-Knudsen (1993), “White Paper, Implementation of Soil/Vegetative Covers for

Final Remediation of the Rocky Mountain Arsenal,” prepared by Morrison-Knudsen
Corporation, Denver, CO, for Shell Oil Company, December.

Othman, M. A., Bonaparte, R. and Gross, B. A. (1997), “Preliminary Results of Study of

Composite Liner Field Performance,” Proc. GRI-10 Conference on “Field Performance
of Geosynthetics and Geosynthetic Related Systems,” R. M. Koerner, G. R. Koerner
and Y. G. Hsuan, Eds., GII Publications, Philadelphia, PA, pp. 115-142. (Final Report
to be published by U. S. EPA, Cincinnati, OH, 2002)

Thornwaite, C. W. (1948), “An Approach Toward a Rational Classification of Climate,”

Geographic Review, Vol. 38, pp. 89-99.

- 12 -