Page 1

RESIDUAL SOILS AS DAM FOUNDATION AND FILL MATERIAL

TABLE OF CONTENTS

1 INTRODUCTION............................................................................................................................................ 3

2 RESIDUAL SOILS .......................................................................................................................................... 5

2.1 ORIGIN...................................................................................................................................................... 5
2.1.1 Weathering Processes........................................................................................................................ 5
2.1.2 Factors affecting weathering processes............................................................................................ 6
2.2 DEFINITION ............................................................................................................................................. 7
2.2.1 Laterites and lateritic soils ................................................................................................................ 7
2.2.2 Saprolites ........................................................................................................................................... 8
3 RESIDUAL SOIL PROFILE .......................................................................................................................... 9

4 MINERALOGY AND STRUCTURE .......................................................................................................... 12

5 CLASSIFICATION, SAMPLING AND TESTING.................................................................................... 14

5.1 CLASSIFICATION ................................................................................................................................. 14
5.2 FIELD IDENTIFICATION...................................................................................................................... 16
5.3 SAMPLING ............................................................................................................................................. 17
5.4 INDEX TESTS......................................................................................................................................... 18
5.4.1 Natural Moisture Content ............................................................................................................... 18
5.4.2 Atterberg limits ................................................................................................................................ 19
6 ENGINEERING PROPERTIES................................................................................................................... 26
6.1 PERMEABILITY..................................................................................................................................... 26
6.2 COMPRESSIBILITY AND SETTLEMENT .......................................................................................... 28
6.2.1 Collapsibility .................................................................................................................................... 28
6.2.2 Compression indexes ....................................................................................................................... 30
6.2.3 Heave ............................................................................................................................................... 31
6.2.4 Compressibility Design considerations ........................................................................................... 32
6.3 COMPACTION PROPERTIES: MAXIMUM DRY DENSITY AND.................................................... 32
OPTIMUM WATER CONTENT ........................................................................................................................ 32
6.4 SHEAR STRENGTH............................................................................................................................... 34
6.4.1 Stress history.................................................................................................................................... 34
6.4.2 Bonding between particles .............................................................................................................. 35
6.4.3 Drained vs. Undrained shear strength............................................................................................ 35
6.4.4 Mineralogy....................................................................................................................................... 36
6.4.5 Relic structures and discontinuities ................................................................................................ 37
6.4.6 Partial Saturation ............................................................................................................................ 37
6.4.7 Determination of shear strength parameters.................................................................................. 38
7 ADDITIONAL DESIGN CONSIDERATION............................................................................................. 39
7.1 LATERITE AND SAPROLITE AS DAM FOUNDATION MATERIAL.............................................. 39
7.1.1 TYPICAL PROBLEMS ..................................................................................................................... 39
7.1.2 FOUNDATION TREATMENT ......................................................................................................... 39
7.1.3 INSTRUMENTATION ...................................................................................................................... 39
7.2 LATERITE AND SAPROLITE AS DAM FILLING MATERIAL ........................................................ 39
7.3 CONSTRUCTION METHODS............................................................................................................... 39
 Page 2

7.3.1 TIME RATE OF CONSTRUCTION ................................................................................................. 39

7.3.2 COMPACTION REQUIREMENTS .................................................................................................. 39
8 PERFORMANCE DURING OPERATION ................................................................................................ 39
8.1.1 PORE WATER PRESSURES............................................................................................................ 39
8.1.2 SEEPAGE......................................................................................................................................... 39
8.1.3 LONG TERM BEHAVIOR................................................................................................................ 39
8.1.4 STABILITY PROBLEMS .................................................................................................................. 39
9 CONCLUSIONS ............................................................................................................................................ 40

10 CASE HISTORIES .................................................................................................................................... 40

10.1 COLOMBIAN DAMS................................................................................................................................... 40
10.2 BRAZILIAN DAMS ..................................................................................................................................... 40
10.3 VENEZUELAN DAMS ................................................................................................................................. 41
10.4 ADDITIONAL DAMS ................................................................................................................................... 41
11 BIBLIOGRAPHY ...................................................................................................................................... 42
 Page 3

1 INTRODUCTION

Most dams being designed or built nowadays are located in developing countries, often situated
in tropical regions.

A crucial component in the design of dams is the knowledge acquired by precedent.

Extrapolation of this precedent in dam construction can only be utilized properly and
successfully through a clear understanding of the conceptual assumptions, local conditions and
construction methods. A clear difficulty to take advantage of precedent in dam projects located
in tropical regions is the presence of a particular type of soils: residual soils.

Unlike the more common transported soils, residual soils originate from the in situ weathering
and decomposition of the parent rock, which has not been transported from its original setting.
The conditions present in humid temperate and tropical climates provide the adequate moisture
and temperature conditions to transform through weathering processes the underlying rocks
into residual soils. Typically these regions have a high potential for hydraulic projects, either
for hydropower, or flood control and irrigation.

Dam projects located in these areas, can encounter thick layers of residual soils at their
foundation. For economical reasons, this material can also be used as fill material. The main
difficulty in dealing with these soils for engineering purposes is that their characteristics are
quite distinctively different from those of transported soils (Blight 1997). Furthermore
concepts and methodologies typically used by geotechnical engineers to deal with transported
soils commonly cannot be utilized with residual soils, or simply are inapplicable. Despite these
inconveniences, the widespread presence of residual soils does not allow us to avoid them, but
challenges us, to deepen our understanding and knowledge about them, to be able to accept and
employ them to our advantage in our projects.

Most of the references in the technical literature about residual soils are written by geologists,
focusing mainly in the description and characterization of these soils. Very limited information
is found related to engineering purposes and specific projects. The father of soil mechanics,
Karl Terzaghi, was also a pioneer in this field. His work on Sasamua Dam (Kenya), in the late
forties and early fifties, opened the possibility of using residual soils in modern dams, through
an understanding of their characteristic behavior. Most of the latter knowledge developed latter
took place in places like South Africa, Brazil, Australia, New Zealand, Colombia and the
United States of America.

The purpose of this bulletin is to produce a state of the art report on “Laterite and Saprolite as
Dam Foundation and Fill Material for Dams”. The intention is not to provide detailed
geological descriptions or characterization methods. The geologic literature offers a plethora of
articles on these issues. The bulletin rather focuses on the dam engineering implications of
dealing with residual soils. The unique aspects of profile descriptions, sampling and testing,
engineering properties, design criteria, construction techniques, and registered behavior is
 Page 4

outlined. An extensive gathering of case histories around the world of dams built on or with
residual soils illustrates these topics in a pragmatic matter.

Hopefully the material included in this report will provide valuable information for dam
engineers to design safer, good performing, more efficient and economical dam structures were
residual soils are present.
 Page 5

2 RESIDUAL SOILS

2.1 ORIGIN

Unlike transported soils, mainly derived from alluvial, lacustrial, marine, aeolic or glacial
depositional processes; residual soils originate in-situ. The formation of residual soils is
controlled by other factors besides than the rock itself. Local climatic conditions and
topography are the most relevant aspects, combined with other variables to induce chemical
(decomposition), physical (disintegration) and biological weathering processes that provide
certain environmental conditions for the development of residual soils from the parent rock
material.

Residual soils are mainly derived from metamorphic and igneous rocks, but residual soils
developed from sedimentary rocks (e.g. shales and loess) are not rare.

2.1.1 Weathering Processes

Weathering results from the reaction of pre-existing materials to different conditions of

pressure, temperature, fluids and stress regimes near the ground surface. Usually physical
weathering precedes chemical weathering. Exogenetic (climate and vegetation) and
endogenetic (structure and composition of the material) determine the extent and the rate of
weathering (Birkeland, 1984). Stress release by erosion, salt crystallization, differential
thermal strain, swelling and ice are the most noticeable physical weathering agents. They
expose fresh surfaces to the chemical agents and increase permeability allowing the flow of
chemically active substances (Blight 1997).

2.1.1.1 Physical Weathering

Physical weathering terminates in the collapse of the parent material and the diminution of its
grain size, as rock disintegrates and stress is exerted along planar structures (bedding, or
fractures) by expansion of the rocks or minerals themselves, or by foreign agents in the voids
(Birkeland, 1984). As depth increases physical weathering is less effective, due to the increase
in confining stresses that inhibits the formation or opening of fractures.
 Page 6

2.1.1.2 Chemical Weathering

Due to constantly changing conditions at the Earth’s surface, rocks and minerals decompose to
reach equilibrium under new environments, altering original materials and generating new
minerals (Birkeland, 1984). Oxidation, reduction, hydrolysis, solution and cation exchange
play an important role in the chemical processes leading to chemical weathering.

Biological weathering comprises both physical and weathering processes. Physical, in the form
of splitting by root wedging, and chemical as bacteriological oxidation.

2.1.2 Factors affecting weathering processes

The parent rock involved is a crucial factor to determine what type of process or agent has a
stronger weathering influence. Physical processes usually control weathering of sedimentary
and metamorphic rocks and chemical processes dominate the weathering of igneous
rocks. Although there might be a predominant process, there is always a presence and
influence of the other weathering factors.

Factors affecting weathering can be divided in macro - factors and micro - factors as follows:

Macro - factors Micro - factors

Climate (moisture, temperature) Lithology
Relief (topography) Structure
Lithology. Goldich’s Series (weathering Fluctuations of water table in recent
susceptibility series) geologic time
Time

Table No. 2.1 – Macro - factors and micro - factors affecting weathering

Among the local variables, climate is the one with the higher influence in the type and rate of
weathering present in a certain region (Blight 1997, Weinert 1974). Physical weathering is
prevalent in dry climates, decreasing in influence as higher moisture and temperature promote
biological and chemical weathering.

Strakhov (1967) proposed a diagram relating the effects of global climate on the parent material
(rock) weathering and the development of clay minerals and other weathering
products. Strakhov’s diagram is shown in Figure No. 2.1. The model does not include local
climatic variations or topographical influences, but nevertheless is a good approximation on the
importance of climate in the weathering processes. It is important to note that even though
general models are valuable, the evaluation of specific local and regional conditions is
fundamental to understand the processes and effects of weathering at a particular site. It can be
seen that at the tropic forest zone, high moisture and temperatures result in an extensive and
rapid weathering and leaching of mobile constituents.
 Page 7

Figure No. 2.1 - Depth of weathering relative to some environmental factors

(After Strakhov, 1967)

Topographic relief and geomorphology affect the depth of weathering, as they control drainage
(both surface flow and groundwater), and consequently the rate of leaching.

The development of residual soil profiles is restricted by erosion removal rate. The weathering
advance must be higher than the removal of weathering products. Residual soil profiles will
generally be encountered in valleys and gentle slopes, rather than on steep slopes
(Blight 1997). Studies (Van der Merwe, 1965 and Fitzpatrick et al, 1977) have shown that
samples taken on high grounds and steep slopes have a greater concentration of kaolinite, while
down slope a higher percentage of montmorillonite (smectites). Landsliding and general mass
movements, promoted by steep slopes also arrest the development of deep residual profiles.

2.2 DEFINITION

For the purposes of this bulletin laterites and saprolites are defined as follows:

2.2.1 Laterites and lateritic soils

Several definitions for Laterites have been proposed, and currently there still no general
consensus. In 1807 Buchanan defined a laterite as a soil with no stratification, presenting large
amounts of iron, forming red or yellow “ocres”, cuttable when wet, but when air dried hard like
a brick. The hardened crusts characteristic of some laterites have associated with a general
 Page 8

group named duricrusts (Fookes et al, 1997). In general terms lateritic soils are yellow and red
tropical soils and concretionary hardened materials, part of the superficial soil horizon (Melfi,
1985). In Deere and Patton’s (1971) weathering profile, laterites and lateritic soils are
generally part of zone IB. (Chapter 2).

Some engineers and geologists define Laterites through low silica-sesquioxides

relation. Sandstones and granites are the most common rocks subject to lateritization.

2.2.2 Saprolites

Saprolites are weak, friable, chemically weathered material, which preserves the original
structure and texture of the parent rock, in spite of the decrease in strength and replacement of
original materials (mainly by clay). Despite having the original structure and fabric of the
parent rock, the mechanical and engineering behavior are very different. Saprolite or saprolitic
soils characteristically present an early stage weathering process.

Their parent rocks are either metamorphic or igneous. In the weathering profile proposed by
Deere and Patton (Chapter 2) correspond to horizon IC and the transition zone II A. In some
cases their in-situ thickness may exceed 10 m, depending on the parent rock material and local
conditions.

For the purposes of this bulletin a material is considered saprolitic if it is soil in the
geotechnical sense.
 Page 9

3 RESIDUAL SOIL PROFILE

Physical-chemical factors affecting weathering are more readily present at the surface.
Therefore the extent of weathering usually decreases with depth. In order to understand the
engineering significance of the gradual transition of the physical properties of materials
subjected to weathering, various authors have proposed typical weathering profiles. Among the
most widely recognized, Deere and Patton’s (1971) model is utilized for purposes of this
bulletin. The profile is subdivided in three major zones: (I) residual soil, (II) weathered rock
and (III) relatively unweathered and fresh rock. Zone I, the residual soil, is further divided in
three, according to the pedologists classification: IA (horizon A), IB (horizon B), IC
(Horizon C). In some instances this model does not fulfill the requirements for an adequate
description, especially if there are pronounced lateral variations. Table No. 3.1 presents some
of the more relevant characteristics of the different zones of the weathering profile.
Figure No. 3.1 illustrates typical weathering for igneous and metamorphic rocks.

Zone Description
IA Horizon Topsoil, roots, organic material, zone of leaching and eluviation. May be
porous. Sandy textures. Medium to high permeability. Low to medium
strength.
I. Residual Soil IB Horizon Characteristically clay enriched; accumulation of Fe, Al, and Si; hence may
be cemented or susceptible to irreversible hardening. No relict structure
present. Deposition of solid materials from IA. Low permeability and
strength (high if cemented).
IC Horizon Relict rock structures retained; silty grading to sandy material; less than
10% corestones. Often micaceous. Most minerals of parent material
(except quartz) altered. Saprolite. Medium permeability. Low to medium
strength.
II. Weathered IIA Transition from saprolite to weathered rock. Highly variable, soil like to
Rock Transition rocklike; matrix medium to coarse sand or silty and micaceous; 10 to 95%
corestones; spheroidal weathering common. Wide range of engineering
properties. High permeability. Relic structures control strength.
RQD 0-50%.
IIB Partly Soft to hard rock; joints stained to altered; some alteration of feldspars and
weathered micas. Lower strength, lower modulus and higher permeability than
rock original material. Medium to high permeability and strength.
RQD 50-75%
III. Unweathered No iron stains to trace along joints; no weathering of feldspars or micas.
rock RQD > 75%.

Table No. 3.1 - Description of weathering profile for igneous and metamorphic rocks
(Adapted from Deere and Patton, 1971)
 Page 10

Figure No. 3.1 - Typical weathering profile for igneous and metamorphic rocks
(Deere and Patton, 1971)

Among the other widely known and recognized models for weathering profiles, the most
referred models are those proposed by Little (1969), Hough (1957).

The Working Party of Geological Society of London (Tropical Residual Soils Report, 1997)
adapted a scale based on different weathering stages, initially proposed by Little (1969),
classifying the weathering grades in six different groups (Figure No. 3.2). This weathering
scale and Deere and Patton’s (1971) weathering profile have many aspects in common, and
both can be applied for engineering purposes.
 Page 11

Figure No. 3.2 - Weathering Grades proposed by the Geological Society of London (1993)

If a sharp basal weathering front is not present, the total depth of weathering may be difficult to
determine, particularly for sedimentary and metamorphic fissile rocks. In fractured crystalline
rocks the task of determining the depth of weathering can be more precise. Fresh outcrops can
evidence an influence 15 to 20 m deep, with occasional abrupt gullies up to 50 m deep.
 Page 12

4 MINERALOGY AND STRUCTURE

Mineralogy is responsible in great part for the wide range of engineering properties exhibited
by residual soils. Their mineralogy is to certain extent inherited from their parent rock
materials and in part due to weathering processes present in their formation. Granites generate
tan and yellow silts and silty sands, with varying quantities of mica and clays of the kaolinite
family. Basalts (rich in ferromagnesian minerals) produce highly plastic montmorillonite clays
with iron oxide coloration varying from deep red to darks browns, typical of “cotton soils”
from India (Sowers, 1970).

Climatological factors involved in the genesis of tropical soils generate geochemical processes
that result in soils with less silica and cations, but richer in iron and aluminum (typical of
saprolites). Laterites or lateritic soils tend to be richer in oxides. As mentioned earlier, laterites
are regarded as having a low silica-sesquioxides relation (Morais Leme, 1985).

Clay minerals are generally found, sometimes as residual minerals from weathering and others
from the parent materials when claystones and mudstones are present
(Fookes et al, 1997). Tropical climate promotes the formation of the clay mineral kaolinite,
making it the most common clay mineral in tropical residual soils
(Queiroz de Carvalho, 1997). Particles of kaolinite are larger and less active than most clay
minerals, but are also platey with a low coefficient of interparticle friction. When shearing
occurs kaolinite orient their longer sides and present slickensided surfaces and low residual
angles of friction (Fookes et al, 1997).

Under certain conditions halloysite is also present, and as a consequence soils containing this
mineral exhibit irreversible changes in index properties if dehydrated. Since tropical residual
soils are frequently found partly saturated, when they are wetted and contain halloysite a
significant decrease in shear strength can be expected (Queiroz de Carvalho, 1997). Unlike
other clay minerals, halloysite and allophane (another mineral present in residual soils) are not
of platey nature and consequently do not exhibit a sharp decrease in their residual
strength. They display low swelling and small strains when wetted or dried
(Fookes et al, 1997). Allophanic soils present satisfactory engineering properties, despite
having liquid limits ranging from about 80 to 250% (Wesley, 1997).

Smectites (montmorillonite) are found in specific types of residual soils (i.e. vertisols or cotton
soils). When their presence is important, small changes in effective stress cause large volume
changes, and when desiccated and wetted result in significant heave displacements
(Fookes et al, 1997).

Table No. 4.1, presents some typical angles of shearing resistance and residual angles of
shearing resistance for different types of minerals found in tropical residual soils:
 Page 13

Clay mineral φ' φ'r

Smectites φ' 15-20 5-11
Kaolinites φ' 22-30 12-18
Allophane φ' 30-40 30-40
Halloysite φ' 25-35 25-35

Table No. 4.1 - Angles of shearing resistance for the different clay minerals present in
residual soils (After Fookes et al, 1997)
 Page 14

5 CLASSIFICATION, SAMPLING AND TESTING

5.1 CLASSIFICATION

The Unified Soil Classification System (USCS) was conceived to classify soils from a
transported origin, showing numerous practical virtues for this type of soils. On the other hand,
when the USCS is applied for residual soils, several disadvantages and problems arise. Typical
engineering properties associated to different classifications groups of the USCS are not
reflected by residual soils.

Several authors have formulated some of the conflicts of “conventional” classification systems
for residual soils (Queiroz de Carvalho 1985-1997, Wesley 1997, Fookes et al-1997):

The physical properties of residual soils are not reproduced successfully in a laboratory
environment.

The unusual clay components that sometimes present in residual soils, generate abnormal
behavior if compared to those associated with traditional classification.

The in-situ characteristics of these materials and their sequence are not adequately described.

The orthodox classification systems are focused on the properties of the soil in a remolded
state, with good results for transported soils. Residual soils are not found naturally in a
remolded state, and this procedure eliminates inherited relic structures of the parent material,
crucial for predicting behavior.

Standard soil classification systems were developed for soils from temperate soils, limiting
their use to the distinctive characteristics of tropical residual soils.

Typical tests (according to ASTM procedures) used for transported soils are not always
applicable, relevant or representative of residual soils characteristics.

Usual relationships and correlations are difficult to establish, and empirical relationships have
to be adapted to local experience.

Frequently residual soils have better engineering properties than index tests suggest, mainly
when index tests results are correlated to the engineering behavior of transported soils as in the
USCS classification system

Due to the aforementioned limitations of traditional classification systems, alternative schemes

have been proposed. The systems used or proposed for residual soils are based on
characteristics sometimes different from grading and typical physical properties, with the
objective of providing some of the information that is lacking in the specific conditions of
 Page 15

residual soils, the characteristics observed at the site and weathering effects should be included.

Considering the difficulties that have been found attempting to develop an adequate
classification system, they have developed basically from two criteria, the pedological and the
physical-chemical characteristics.

In any case, when analyzing soil profiles including residual soils, and their classification is
either based on the Unified Soils Classification System, it should be clear to the engineer which
of the properties can be different under real conditions. Some of the classification system that
have been proposed are:

Wesley (1997), presented a classification system to complement and use in conjunction with
the Unified Soil Classification System (USCS), including an analysis of the mineral
components, and based on the following points:

Soil composition, including mineralogy and physical properties of the soils.

Structure in its undisturbed in- situ state, both macro-structure and micro-structure.

The proposed system by Wesley classifies soils in three main groups that tend to exhibit similar
engineering properties:

Group A: Residual soils without a strong mineralogical influence.

Group B: Residual soils with a strong mineralogical influence deriving from commonly
occurring clay mineral.

Group C: Residual soils with a strong mineralogical influence deriving from special clay
minerals only found in residual soils.

The three main categories are further subdivided into sub-groups to limit the variables affecting
engineering behavior. As pointed by Wesley, one of the main difficulties applying this method
is its use of mineralogical composition as a starting point for classification. On the other hand
the scheme has the advantage of including most residual soils, denoting simplicity and
applicability.

A summary of this classification system and the characteristics of each group is illustrated in
Tables No. 3.1 and No. 4.1.

The Geological Society suggests the use of the formal French Classification System for residual
soils. Although this method of classification is intended for engineering purposes, anticipating
typical behaviour, the terminology is more complex and less practical than Wesley’s System.

Several other classification systems have been proposed. They can be grouped in three main
categories (Queiroz de Carvalho (1985):
 Page 16

Classifications based on chemical, pedological and morphological criteria (Clare’s - 1957,

based on the silica/ alumina and silica/ sesquioxides ratios- 1951, Ruddock’s - 1969, Little’s –
1967, Gidigasu’s – 1971).

Orthodox Classifications (USCS, Highway Research Board, Vallerga - 1969, Lal and Bindra’s
- 1981, Eklu - Natei and Muller’s - 1981, Medina and Preussler’s - 1980 and Vargas - 1982).

Classifications based on non-orthodox tests (De Graft Johhnson, Arulanandan, 1969, Queiroz
de Carvalho’s - 1981, Lohnes and Demirel - 1973, Tuncer’s - 1976, Nogami and Villibor -
1981, EPF Zurich- 1979).

Every abovementioned classification system has specific benefits, but it is believed the use of
Wesley’s scheme in combination with the USCS provides the most complete and practical
classification system. For that reason the other classification systems are only referenced and
not explained in detail.

5.2 FIELD IDENTIFICATION

Due to the sensitivity of residual soils, an adequate field classification is fundamental to

properly classify and correlate engineering properties of residual soils.

The identification of saprolite and the other residual soils shall be mainly performed in the
field, Test holes and wide borings are the most suitable means for establishing the soil and
weathering profile. Since most residual soils are cohesive and frequently the water table is
found deeper than most of the layers, the holes tend to be stable and provide time for an
adequate description (Simmons, 1997).

The information that should be collected in the field includes the following aspects (Fookes et
al- 1997, Simmons-1997):

• Origin (rock type)

• Texture
• Fabric or structure
• Colour
• Resistance
• Natural Moisture
• Density
• Apparent behavior
• Mineralogy
• Strength
 Page 17

In addition, factors like precise location, landforms characteristic of the site, climate and
vegetation should be recorded, since they can give a hint to the extent and variation of
weathering.

In most cases where saprolitic soils are encountered, engineering behavior is controlled by
inherited geological features from the parent material. As a result it is very important to detect
and describe any preferential planes, joints, or any other features remaining from the parent
rock.

Conventional geophysical methods, may encounter problems, since it is normal that residual
soils in higher layers (highly weathered), have higher wave velocities, making a correct
interpretation of the refraction tests difficult. Occasionally cross-hole tests have been used, but
they also present a number of restrains, including complications to determine layer
thickness. In any case, if any of these tests is used, it should be calibrated and verified with
profiles inferred from drill- holes.

5.3 SAMPLING

This section emphasizes the special sampling practices and specific criteria applicable to
residual soils. If not specifically commented or disapproved in this chapter, typical sampling
techniques for transported soils can be applied.

Two main matters are fundamental for an adequate sampling program: representativity and
disturbance. Typical anisotropy, heterogeneity and variability of tropical residual soils
complicates the planning of a sampling agenda. Sampling for testing of many engineering
properties (i.e strength) requires large specimens to ensure that variables like structure and relic
features are representative of the mass. The weak cementation and bonding of particles of
residual soil also sets further hurdles, because the common brittle structure of this type of soils
makes them very sensitive to disturbance (Fookes et al, 1997).

The large voids commonly present in residual soils and the fact they are frequently found partly
saturated generates inevitable volumetric expansion when sampled, especially in saprolites,
where the stress changes are greater (Fookes et al, 1997).

Depending on the type of engineering property the sample is going to be tested for,
requirements vary. Disturbed samples can be utilized for tests related to density, water content,
particles size, and compaction. For this cases, SPT samples, open-drive tube samples or piston
samples are satisfactory. The assessment of strength and stiffness in the laboratory requires
high quality samples; hand cut samples from trial pits or shafts. The importance of this aspect
has been reported by Prusza (1984) at the Guri Dam (Venezuela), where he noticed important
differences between shear tests performed on samples obtained by Denison samplers and
undisturbed blocks.
 Page 18

Generally, special care should be taken to preserve the sample’s in-situ water content,
especially when the soil exhibits the minerals halloysite and allophane, since irreversible
changes can occur if the material is allowed to desiccate.

Wolle et al, separate residual soil in three groups according to the difficulty of sampling.
Lateritic soils are considered easy to sample, because of their relative homogeneity, less
sensitivity to disturbance. These conditions make them more representative of the in-situ
conditions. On other hand, saprolites are regarded as difficult to sample. Relic structures,
boulders and sensitivity to disturbance can even inhibit any sampling, particularly in the soil-
rock transition. In general terms, with increasing depth greater heterogeneity id found, and
consequently samples are less representative and difficult to extract.

Shelby tube samplers have been more commonly used than Denison samplers, but even this
type of sampler becomes unfeasible to use in saprolitic soils when resistance increases.

5.4 INDEX TESTS

In spite of the fact that traditional index test do not properly typify residual soils, their extensive
use for geotechnical analysis worldwide has promoted different studies to adapt these tests
(or their results) to the different behavior and characteristics of residual soils, in order to allow
for better interpretations and correlations of their engineering properties, base on index tests.

A basic principle in geotechnical testing is to reproduce in-situ conditions in the laboratory to

anticipate the behavior of soils for design purposes. This fundamental requirement for useful
and successful testing in usually not satisfied for residual soils. As de Mello (1972) pointed
out, only their behavior after complete disaggregation, remoulding and saturation
is modeled. Furthermore, these soils neither have been, nor will be subjected in-situ to these
conditions.

The single individual factor that has a greater effect on index properties is drying. Due to the
mineralogical composition of residual soils, drying can induce irreversible chemical changes,
leading to alteration of the clay minerals and aggregation of fine particles
(Fookes et al, 1997). This circumstance can have profound effects in the index properties of
residual soils, that is in their moisture content, plasticity, shrinkage and particle size
distribution. The main effects of drying are increased cementation due oxidation of the iron
and aluminum sesquioxides and dehydration of halloysite and allophane minerals
(Townsend, 1895).

5.4.1 Natural Moisture Content

Fourie et al (1997) and Fookes et al (1997) recommend a specific methodology for determining
the natural moisture content of residual soils, to subtract any structural water present in the soil,
form the actual moisture content.
 Page 19

Two samples should be used to determine natural moisture.

One of them should be dried to a standard temperature of 105 degrees Celsius, calculating it’s
natural moisture with the conventional method.

The second sample should be dried either with air, or oven-dried at a temperature of 50 degrees
Celsius and a relative humidity of 30%.

If a significant difference is observed, it indicates that structural water is present within the soil
sample. If such difference exists all the other test samples that need moisture content
measurement should use the second procedure.

The dams built on the “Antioqueño” Batholith residual soils in Colombia (Piedras Blancas,
Quebradona, Troneras, Miraflores, Santa Rita I and II, Punchiná and San Lorenzo) exhibited
high water contents, ranging between 25-40%.

5.4.2 Atterberg limits

To avoid alteration of natural properties, the determination of Atterberg limits should be

performed without any form of a priori drying. The Atterberg limits, as well as the natural
water content, should be measured with the fraction of the soil passing the 425 m sieve
(Fookes et al- 1997, Fourie et al-1997). Morin and Todor (1975) gathered reported effects of
drying on the plasticity of volcanic soils in different parts of the world (Figure No. 5.1).

Figure No. 5.1 - Effect of drying on the plasticity of volcanic soils in different parts of the
world (Morin & Todor, 1975)

The effects of drying on the Atterberg limits are also showed in table # for soils around the
world with different mineralogy (Fookes et al, 1997). Values are given for air-dried and oven
 Page 20

dried samples. Notice that the effect is minor for soils not containing allophane and halloysite.

Atterberg limits
Soil location oven
and type air dried
natural WL:WP dried
WL:WP
WL:WP
Costa Rica:
Laterite 81:29 56:19
Andosol 92:67 66.47
Dominica:
Allophane 101:69 56:43
Latosolic 93:56 71:43
Smectoid 68:25 47:21
Hawaii:
Humic latosol 164:162 93:89
Hydrol latosol 206:192 61:NP
Java:
Andosol 184:146 80:74
Kenya:
Red clay, Sasumua 101:70 77: 61 65:47
Malaysia:
Weathered shale 56:24 48:24 47:23
Weathered granite 77:42 71:42 68:37
Weathered basalt 115:50 91:49 69:49
New Guinea:
Andosol 145:75 NP
Vanuatu:
Volcanic ash,
261:184 192:121 NP
Pentecost
Note: NP indicates non-plastic

Table 5.1 - Effects of drying on the Atterberg Limits of residual soils in different parts of
the world (Fookes et al, 1997)

One of the limitations using conventional Atterberg limits tests to classify residual soils and
correlate with engineering properties of soils with similar values is the destruction of the
cementing bonds between the particles.

Pre-drying and mixing in standard tests, has been shown to have important influence on results
(Townsend –1985, Fookes et al- 1997, Fourie- 1997). Some recommendations proposed to
diminish these effects are:
 Page 21

Control temperature, and avoid sensible changes form in-situ conditions.

Mixing time should be limited to a maximum of 5 minutes, using fresh samples for each of the
test points, this practice avoids an increment of liquid limit and plasticity index in some
residual soils due to the breaking down of particles after prolonged periods of mixing.

The sample should not be dried for seeming; the soil should be broken down using distilled
water. If distilled water produces property changes caused by soluble salts, a soil water
solution should be used.

Mineralogy has also a strong effect on the Atterberg limits. Residual soils containing particular
prevailing minerals tend to behave as a group with regards to Atterberg limits, exhibiting values
in a relative close range. When minerals like montmorillonite, allophane and halloysite are
present, the aforementioned behavior is applicable. Wesley et al (1997) plotted residual soils
containing these minerals in plasticity chart, where typical regions in for soils containing
different minerals can be identified (Figure 5.2).

Figure 5.2 - Influence of mineralogy on the position on the plasticity chart for residual
soils (Wesley et al, 1997)

In the classification system proposed by Vargas (1982) proposed that the plasticity chart be
plotted aside from the activity for residual soils. The idea behind this suggestion is to identify
the residual soils that are affected by manipulation (Melfi et al, 1985). It should be noted that
silty micaceous and / or kaolinitic saprolitic soil exhibit high activity, despite having inactive
minerals of the kaolinitic group in their clay fraction. This behavior has been attributed to
micaceous or kaolinitic silt, which increases plasticity (Melfi et al, 1985).
 Page 22

Figure No. 5.3 - Plasticity chart associated with activity for residual soils (Vargas, 1992)

The dams built on the “Antioqueño” Batholith (Colombia) residual soils (Piedras Blancas,
Quebradona, Troneras, Miraflores, Santa Rita I and II, Punchiná and San Lorenzo) presented
liquid limits (wL) between 40 to 50% and plasticity indexes high water between 7-12%.

Figure No. 5.4 through 5.7 present liquid limit (wL) and plasticity index (IP) values for clayey
lateritic soils and saprolitic soils for Brazilian Dam sites.

80
Liquid Limit wl (%)

60
40
20
0
Itaparica Três Juquía I Juquía II Poços de Itaipu
Irmãos Caldas

Figure No. 5.4 - Liquid limit (wL) values of lateritic soils for Brazilian Dam sites
(Texeira et al, 1985)
 Page 23

100

Liquid Limits wl (%)

80
60
40
20
0
Nova Euclides da Passauna Emborcação
Avanhandava Cunha

Figure No. 5.5 - Liquid limit (wL) values of saprolitic soils for Brazilian Dam sites
(Texeira et al, 1985)
Plasticity Index IP (%)

50
40
30
20
10
0
Itaparica Três Juquía I Juquía II Poços de Itaipu
Irmãos Caldas

Figure No. 5.6 - Plasticity Index (IP) values of lateritic soils for Brazilian Dam sites
(Texeira et al, 1985)
Plasticity Index IP

40
30
20
(%)

10
0
Nova Euclides da Passauna Emborcação
Avanhandava Cunha

Figure No. 5.7 - Plasticity Index (IP) values of saprolitic soils at Brazilian Dam sites
(Texeira et al, 1985)

5.4.2.1 Specific Gravity

Residual soils can exhibit unusual low specific gravity value, due to the mineralogy (i.e.
allophane). The soil should not be dried before performing this test, and the specific gravity
should be established at the soils natural moisture content. (Fookes et al - 1997, Fourie - 1997)
Figure No. 5.8 shows typical Specific Gravity values for clayey lateritic soils and saprolitic
soils for Brazilian Dam sites.
 Page 24

Specific Gravity Gs
3.1
3
2.9
2.8
2.7
2.6
2.5
2.4

ha
u

av
ip

ric

un
Ita

nd
pa

C
Ita

ha

da
an

es
Av

id
a

cl
ov

Eu
N
Figure No. 5.8 - Specific Gravity values of lateritic and saprolitic soils at Brazilian Dam
sites (Texeira et al, 1985)

5.4.2.2 Particle size distribution

For determining the particle size distribution of soil sample, drying should also be avoided, and
the soil should we weighted at its natural moisture content. Instead of drying the sample, a
dispersant solution can be utilized to sieve the sample in a wet state. Chemical treatment
should not be used, but a dispersants is recommend prior to sedimentation.
(Fookes et al, 1997)

Figure No. 5.9 and 5.10 present the percentage of sand and fine material for clayey lateritic
soils and saprolitic soils for Brazilian Dam sites.

Figure 5.9 shows the particle size distribution for the residual soils of Guri Dam in Venezuela
(Chaudry, 1979). Remark, that the percentage of clay decreases with depth (especially when
the saprolite is encountered), while the percentage of coarse particles (i.e. sands and gravels)
increases.
90
80
70
% of total mass

60
50
40
30
20
10
0
Itaparica Três Irmãos Juquía I Juquía II Poços de Itaipu
Caldas

% Clay < 2m % Sand

Figure No. 5.9 - Percentage (%) of sand and clay of lateritic soils for Brazilian Dam sites
(Texeira et al, 1985).
 Page 25

100
90
80

% of total mass
70
60
50
40
30
20
10
0
Itaparica Nova Euclides da Passauna Emborcação
Avanhandava Cunha
% Clay < 2m % Sand

Figure No. 5.10 - Percentage (%) of sand and clay of saprolitic soils for Brazilian Dam
sites (Texeira et al, 1985)
 Page 26

6 ENGINEERING PROPERTIES

6.1 PERMEABILITY

As mentioned before a general characteristic of residual soils is the large variability of

engineering properties both vertically and horizontally in the weathering profile. Permeability
is not an exception to this general rule.

More important than in transported soils, it is also essential to differentiate between the
permeability coefficients of residual soils in-situ in their natural state and as fill materials, after
being remolded and compacted. The macrostructure of residual soils generally controls the
permeability. In saprolites, relic structures (joints, bedding planes, quartz veins) are
predominant paths of flow for water seepage (Carga, 1997). Another factor affecting
permeability of tropical residual soils is their cemented structure, that creates big particles and
voids. These features allow for greater permeabilities than would be typical of transported soils
with similar grain size distribution. Consequently, it is evident that using common correlations
for coefficients of permeability based on grading (i.e. Hazen) can be misleading.

Generally, more mature residual soils, close to the surface (i.e. laterites) tend to have relative
low permeabilities, while young residual soils or saprolites exhibit relative permeabilities from
medium to high. Therefore it can be evidenced, that commonly permeabilities increase with
depth in the weathering profile, until sound rock is found, and low permeabilities are
encountered. Superficially, residual soils can be cracked, allowing for an easier penetration of
water and greater permeability (Fookes et al et al, 1997). Laterally, variations of permeability
can also be important, and therefore the use of “typical” values of permeability can be
problematic and misleading (Carga et al, 1997).

Understanding that macro-scale characteristics dominate flow, the question of representativity

of permeability testing arises. Small laboratory tests do not model these type of conditions, due
to the limitations of scale. Therefore permeability field tests are more appropriate to model the
flow properties of residual soils. Since residual soils are mostly found in places where the
water table is relatively deep, pumping tests are sometimes not applicable, and infiltration or
tests have to be used (Carga et al, 1997). The specific choice between the two tests depends on
specific site conditions.

Insects, mainly termites, have generated in some deposits large diameter holes, which can act as
preferential paths of flow. Besides, high seepage flow, this condition can lead to piping under a
dam foundation. This feature has been evidence in the Amazon Region of Brazil, where this
specific type of soils have been named “Canaliculi”. The tubes or galleries have found to range
from a few millimeters to 20 centimeters. De Mello et al (1988) described problems with
termite channels in the foundations of a 30 m high earth dam. “Canaliculi” or biologically
worked soils have been encountered at the following dam sites: Tucuruí, Vereda Grande,
Balbina, Samuel and Kararao (De Morais et al, 1985).
 Page 27

Villegas (1990) reported coefficients of permeability after compaction in the range of 10-7 to
10 –8 m/s for the fill materials for the Troneras and Miraflores Dams. These materials are
mainly silts, sandy silts and silty sands, derived from the weathering of quartzdiorites and
granodiorites (Antioquia Batholith, Colombia). The horizon IB (according to Deere and
Patton’s Classification) was reported with a permeability of 10 –7 m/s, and horizon IC with
10 –6 m/s. According to Villegas the permeability of the saprolite was not much higher than the
lateritic horizon, because the spaces between the boulders of rock where completely filled with
a matrix of weathered material, composed of sand and some silts, that inhibit the flow of water.

Costa Filho et al (1985) presented typical permeability values for residual soils of Brazilian
dams foundations:

Rock Type Residual Soil Permeability (m/s) Type of test

-5 -6
Basalt Mature residual and saprolitic 9,510 – 3 x10 Variable head in piezometers and
infiltration in boreholes
Basalt Mature residual and saprolitic 9,5x10-7 – 9,510-10 Infiltration in boreholes and pits.
Gneiss Mature residual (porous clay) 4,8x10-6 – 9,5x10-7 Infiltration and pumping in
and saprolitic boreholes
Gneiss Mature residual (porous clay) 4,8x10-5 Infiltration and pumping in
and saprolitic boreholes
Gneiss Mature residual (porous clay) 2,2x10-6 Infiltration in pit.
and saprolitic
Gneiss Saprolitic 9,5x10-7 Infiltration in pit and pumping
Gneiss Saprolitic 9,5x10-7 Variable head in lab. permeater
Migmatite Saprolitic 3,2x10-5– 9,5x10-8 Infiltration in boreholes

Table No. 6.1 - Typical permeability values for residual soils of Brazilian dams
foundations (Costa Filho et al, 1985)

In general terms, saprolitic soils derived from granite can exhibit permeability values ranging
from 4 x 10-3 to 5 x 10-9 m/s, while saprolitic soils derived from gneiss have permeability values
ranging from 5x 10-6 to 1 x 10-7 m/s. Mature residual soils originated from the same parent
materials have permeability coefficients from 4x 10-6 to 5 x 10-9 and 5x 10-5 to 1 x 10-6 m/s
respectively (Costa Filho et al, 1985).

The collapsible behavior that some residual soils exhibit (i.e. laterites) due to their fragile
natural structure, can have significant effects on their permeability. When these materials are
wetted major collapse can occur. After collapse takes place, the structure is destroyed,
reducing the void ratio considerably. Consequently, the permeability is also reduced. Itaipú
and Itumbiara dams experienced collapse in their foundation while being impounded (De
Morais et al, 1985). Guri Dam was also founded on collapsible soils (Prusza. 1983). The
collapsibility of a soil is also a function of the applied pressure. For Guri Dam, laboratory tests
indicated permeability decreased from 2x 10-5 to 5 x 10-7 m/s at 50 kPa to 600 kPa respectively
(Prusza et al, 1983).

One of the consequences of the relatively high permeability displayed by residual soils is that
loading and unloading at typical engineering rates rarely results in undrained behavior. This
 Page 28

condition excludes soils that have collapsed and have experienced a void ratio reduction, and as
a consequence can generate excess pore pressures (Fookes et al, 1997).

De Mello at al (1985) recommend the use of two elements to deal with foundation seepages in
residual soils: the use of a cut-off across the pervious horizon and the increase of the seepage
path by utilizing an internal impervious blanket. On the other hand, they discredit the use of
external blankets. Additionally, they suggest the control of the rate of rise during the first
filling, in order to detect foundation seepages and treat them properly. These practices
diminish the hydraulic gradients that can lead to piping and erosion.

6.2 COMPRESSIBILITY AND SETTLEMENT

Like other engineering properties, compressibility is also affected by the considerable

heterogeneity of residual soils along the weathering profile. Compression indexes vary
significantly in the different horizons of the weathering profile. For the purpose of design
settlement estimates, the represantivity of values used has to be analyzed with utmost care. An
evident differentiation between properties of dissimilar layers has to be made.

6.2.1 Collapsibility

A clear understanding of the structure and inter-particle bonding of residual soils is deemed
fundamental to comprehend the compressible behavior of residual soils. The structure of
residual soil is a product of the weathering processes by which they were formed (Fookes et al,
1997). As an effect of leaching and loss of material, residual soils turns into silty or clayey
sand with high void ratio and an unstable collapsible particle structure (Barksdale et al, 1997).

One of the consequences of residual soils structure is that they exhibit a yield stress. In other
words, when a certain state of stress is exceeded they present a discontinuity in their stress-
strain behavior, decreasing in stiffness. In certain types of residual soils after the yield stress is
exceeded, the structure collapses (Fookes et al, 1997). The collapsibility of residual soils is
analogous to a preconsolidated behavior in transported soils. In deed, the yield stress is
sometimes referred as the “quasi-pre-consolidation”. After the yield stress has been exceeded,
a true consolidation stress can be determined, and the compressibility of residual soils can be
analyzed in the same matter as for transported soils. The equivalent preconsolidation pressure
can be a measure of the inter-particle or inter-mineral bonds after weathering
Barksdale et al, 1997). After it is exceeded the bonds are gradually shattered. The collapse
usually occurs when the loaded soils are wetted, eliminating the suctions in the soil, thus
reducing resistance. The collapsible behavior is most important in laterites, since they have
been more strongly weathered. Saprolites do not show this characteristic as often.

Since the settlement suffered by in-situ residual soils is small, if the yield stress is not
exceeded, determining this threshold value is valuable. Yield can only be evidenced if the soil
exhibits a discontinuity in a linear stress - strain plot (Fookes et al, 1997). The use of log-log
plots of the same variables can facilitate the identification of the yield stress
(Vaughan, 1985). Typical void ratio or strains vs. the logarithm of stress curves can inhibit
 Page 29

recognizing the yield point easily (Vaughan, 1985). Fookes et al (1997) summarized the
oedometer data gathered by Vargas (1973) illustrating the discontinuity in stress-strain behavior
after the yield stress is exceeded.

Figure No. 6.1 - Yield stress for residual soils (Vargas, 1973)

As mentioned in the previous chapter, Itaipú and Itumbiara dams experienced collapse in their
foundation while being impounded (De Morais et al, 1985). In Guri Dam, the foundation was
prewetted, by excavating contoured canals at different elevations and filling them with water.
This practice has been followed in places where the water table was found deep, precluding the
collapse of the soils (Prusza. 1983).

Besides the aforementioned measure of prewetting the residual soils to treat residual soils
against collapse, compaction has also been effective. The difficulty with this technique has
been determining the depth of potentially collapsible soil. Among the different type
densification techniques that have been used, the following can be listed: vibrating smooth
wheeled rollers and impact rollers (with or without prior watering) and dynamic compaction
(Barksdale et al, 1997).

Laboratory and field tests can be used to determine the yield stress. In both cases, sample
disturbance and stress path follow during the test, influence the result. Among the reported
tests commonly used are: plate load test, screw plate test, pressuremeter tests, oedometer and
triaxial tests (Barksdale et al, 1997).

Characteristics of Guri Soils compressibility figures.

 Page 30

6.2.2 Compression indexes

As mentioned earlier, residual soils not subjected to disturbance tend to behave as if

overconsolidated. Despite this fact, the at rest coefficient Ko has been found to be in some
non-expansive materials less than unity (approximately 0,5). This condition has been attributed
to the leaching of materials (Barksdale et al, 1997).

The compression index (the slope of the void ratio vs. the logarithm of stress curve) for residual
soils, unlike for transported soils, is a function of the yield stress and the initial
void ratio. Fookes et al (1997) gathered data from several authors correlating the compression
index Ccs and the initial void ratio. The data is presented in Figure No. 6.2.

Figure No. 6.2 - Relationship between compression index and initial void ratio for
residual soils (After Fookes et al, 1997).

Blight & Brummer (1980) performed tests correlating compression indices with different
variables. The results suggest a direct correlation between initial void ratio (eo) and
compression and rebound indices. Figure No. 6.3 shows the relationship for residual weathered
andesite lava.
 Page 31

Figure No. 6.3- Correlation between initial void ratio (eo) and compression and rebound
indices (After Blight & Brummer,1980)

The difficulty of predicting settlements in residual soils can be illustrated with the experience at
Guri Dam. Oedometer tests performed in the laboratory predicted settlements between 4 to14
%, while the actual field measured settlements under a load of about 0,3 MPa was in the range
of 0.4 to 1.2%. Table No. 6.2 illustrates the considerable difference between predicted and
actual settlement values at Guri Dam (Prusza, 1989).

% Settlement at 7 kg/cm2
load
Data From Foundation Fill
Oedometer tests 4,0 - 14 5,0 - 12
Triaxial Consolidation (Ko) test 1,5 – 4,0 -

Triaxial Anisotropic Consolidation (Ko = 0,5) test 1, 5 - 7,0 1,5 - 3,0

Field Instrumentation 1,5 – 3,0 1,2 – 2,5

Table No. 6.2 - Comparison between predicted and actual settlement values at Guri Dam
(After Prusza, 1989)

6.2.3 Heave

Heaving of residual soils usually takes place in semi-arid regions, triggered by seasonal
changes of moisture in the soil. This phenomenon is mostly present in soils having smectite
minerals (montmorillonite), which tend to swell with moisture increases. Residual soils
derived from weathered shales, mudrocks and basic igneous rocks are more prone to swell
(Barksdale et al, 1997).

The reason for heaving due increases in water content, can be explained with a reduction of
 Page 32

suction forces in the soil mass. Usually, the consequences of heaving tend be more notorious in
light structures (i.e. houses). Dams, being massive structures have fewer propensities to be
affected.

6.2.4 Compressibility Design considerations

The heterogeneity of residual soils, leads to dissimilar compressibility characteristics at a dam

site. This circumstance can cause differential settlements in the dam structure. As a result the
dam has to be able to support this type of movements. Experience has shown that the type of
dams most suitable for this behavior are earth core rockfill, homogeneous and concrete face
rockfills.

If a rockfill with central core dam is chosen, it is important to guarantee that the core settles less
than the adjoining shells, avoiding a core-hangup that can generate hydraulic fractures
(De Mello et al, 1985).

6.3 COMPACTION PROPERTIES: MAXIMUM DRY DENSITY AND

OPTIMUM WATER CONTENT

Residual soils have been extensively used as fill materials. They can be very appropriate for a
central core, due to their low permeability after compaction. Residual soils also have been
commonly utilized in homogeneous dams, providing adequate shear strength for the slopes of
the dam body and a watertight barrier.

The high variability of tropical residual soils complicates the control of the compaction
properties of a fill. Obtaining representative samples along a weathering profile can be
misleading. Also, controlling water content in tropical residual soils during placement can
prove to a difficult task, since dam sites located in this areas present high precipitation. This
issue is further explained later in this chapter with some case histories of Colombian dams.
Mineralogy of residual soils also presents additional challenges, since soils containing
halloysite and smectites can present inadequate strength for compaction or can exhibit
intolerable volume changes after hydrating (Simmons et al, 1997). Despite this limitation, soils
containing the aforementioned mineralogy have been used as impervious layers in Sasamua and
Arenal Dams (Simmons et al, 1997).

The compaction characteristics of residual soils should be determined without prior drying,
especially when drying sensitive minerals are present (i.e. allophane and halloysite). The
optimum water content is not the only variable affected by drying; the maximum dry density is
also disturbed. Simmons et al (1997) show considerable effects on these two variables in a
lateritic soil in Ghana (Figure No. 6.4).
 Page 33

Figure No. 6.4 – Sample preparation and drying effects on compaction characteristics of a
lateritic soils in Ghana (After Simmons et al, 1997)

Compaction energy can also influence the compaction characteristics of residual soils. For
instance, this behavior can take place in lateritic soils presenting concretions. The compaction
effort can break down the bonds and cementation between particles, increasing the fine fraction
of the soil. To diminish this problem, it is recommended to utilize fresh samples for every
point to determine the compaction curve (Simmons et al, 1997).

In order to achieve a better representativity of the anticipated compaction characteristics of the

soil, test embankments are recommended to establish the compaction properties of the fill
materials in the field. At Guri Dam trials at the test embankments showed that the sheepfoot
rollers were most appropriate. It was also evidenced that the residual soils derived from gneiss
found at the site should be compacted in layers 0,2 m thick at the most with six passes of the
roller to achieve 96% of the maximum dry density, measured with the Hilf rapid compaction
control method. In general terms, for this case, the Hilf method gave a higher compaction
percentage compared with the Standard Proctor Test (Chaudry et al, 1979). The difference in
values has been attributed to the effect of drying in the Standard Proctor Test that dehydrates
the halloysite and sesquioxides present in the soil. If the sample is not dried before executing
the Standard Proctor Test results tend to be similar for both methods (Chaudry et al, 1979).

Sheepfoot rollers for compaction of the fill at Guri Dam were very effective, always reaching a
compaction percentage higher than 96% and on average higher than 98%. This was
accomplished if the water content was kept within 2% of the optimum (Chaudry et al, 1979).

At Punchiná Dam (Colombia) Villegas (1982) reports the following material average
characteristics after compaction:
 Page 34

Characteristics Average value

Maximum dry density (t/m3) 1,65
Optimum water content (%) 20,1
Field dry density as % of max. γd (t/m3) 95,4

Table No. 6.3 - Punchiná Dam, compaction characteristics (Villegas, 1982)

The soils used for construction of the fill material at Punchiná were mainly sandy silts
classified as SM in the USCS (Villegas, 1982). The placement of the fill at Punchiná Dam was
simplified because the natural and optimum water contents differed only 4.1% on average. The
high precipitations of the area allowed only for a short construction season and consequently
the contractor was allowed to place the material with the same water content as exploited
(Villegas, 1982).

6.4 SHEAR STRENGTH

The analysis of shear strength behavior of tropical residual soils demands taking some unique
factors into consideration that are not as relevant for transported soils. Furthermore, the shear
strength characteristic of saprolitic soils is very different compared to the behavior of laterites.
In the first case, the shear strength characteristics and controlling factors for saprolites resemble
that of rocks because of the presence of the structure and texture of the parent rock.

The most relevant factors influencing shear strength of residual soils are the following (Brenner
et al-1997, Fookes et al, 1997):

• Stress history.
• Bonding between particles.
• Mineralogy.
• Relic structures and discontinuities.
• Anisotropy.
• Widely variable void ratio.
• Partial saturation.

6.4.1 Stress history

Unlike transported soils, stress history does not have an important effect in the shear strength of
residual soils. Weathering progressively modifies the state of stresses that is followed by and
adjustment of the weathered material to reach a state of equilibrium with the prevailing
conditions (Brenner et al, 1997).
 Page 35

6.4.2 Bonding between particles

Leaching and solution of cementing agents generate bonding between the soil mass particles.
The bonds tend to be fragile causing a brittle shear strength behavior. Moreover, inadequate
sampling methods destroy the bonds, and as a consequence shear strength is underestimated.
The bonds tend to have a greater influence on the behavior of a large mass in situ and decrease
with a progressive weathering (Brenner et al, 1997). Following this line of reasoning, small
laboratory tests do not properly reflect the benefit of bonding in shear strength. Brenner et al
(1997) report that in partly weathered rock bonds provide high strength (greater than 200 kPa),
while in mature residual soils the bonds tend to much weaker (lower than 100 kPa).

As mentioned in chapter 6.2 (Compressibility and settlement) interparticle bonds are very
sensitive to saturation, and can be destroyed if the soil is stressed during saturation (Brenner et
al, 1997).

6.4.3 Drained vs. Undrained shear strength

The adequate planning of a laboratory testing program to estimate shear strength requires
anticipating what type of shearing behavior the soil will present: drained or undrained. Residual
soils in-situ usually present a drained behavior, while compacted materials composed of typical
residual soils behave in an undrained matter.

The weathering of residual soils generates leaching of the material, bio-channels and large void
ratios. Also, in-situ saprolitic and lateritic soils commonly behave as if overconsolidated
(Brenner et al, 1997). To further comprehend the shearing behavior of in-situ residual soils
Skempton’s pore pressure parameters can be analyzed. Since they behave as if
overconsolidated, the A parameter (ratio between deviator stress and excess pore pressure
during shear) is usually close to zero at failure. In other words, pore pressures produced during
shearing are usually small and unimportant for analytical purposes (Brenner et al, 1997). In
Volta Grande (≈50 m) and São Simão (≈100 m) dams bearing capacity of the foundation was
estimated based on UU triaxial tests, but concerns about the actual resistance promoted the
installation of piezometers. These instruments did not register any perceptible pore pressures
during construction. The drained behavior during shearing has been attributed to the presence
of macro-pores and air-pores that inhibit the formation of pore pressures (De Mello et al, 1985).

On the other hand, compaction of residual soils destroys the aforementioned features that
promote a drained behavior. Furthermore, the amount of fines inhibits drainage. Several case
histories of dams experiencing high pore pressures in their residual soil fills have been reported.
The compacted material normally presents negative pore water pressures and thus shear
strength is reduced. For these circumstances short-term conditions are the most critical. Hence,
slope stability problems can occur during construction and have to be closely monitored. This
situation was experienced with the dams built on the Colombian “Antioqueño” Batholith
residual soils (Piedras Blancas, Quebradona, Troneras, Miraflores, Santa Rita I and II, Punchiná
and San Lorenzo). The region where these dams are constructed presents a high precipitation
average (2100-5300 mm) with only a short dry season of 3 months during the year. The only
 Page 36

easily available materials are residual soils with water content 2 to 10% above optimum
(Villegas, 1984). Under these difficult circumstances, almost all the available materials were
placed at their natural water content (without prior treatment) at a high rate to take advantage of
the short construction season. This procedure leads to high pore pressures and excessive strains
(Villegas, 1984). The monitoring of these pressures with piezometers was fundamental to avoid
possible shear strength failures and the construction sequence was adapted according to the
observed behavior. Long term monitoring of the pore water pressures during reservoir
operation was difficult, because the large settlements damaged a high percentage of the
instruments.

6.4.4 Mineralogy

The general characteristics of minerals present in residual soils were thoroughly described in
chapter 4. Regarding shear strength, soils containing smectites are the most problematic. This
platey clay minerals exhibit low angles of friction. Additionally, displacements orient these
minerals reducing the shear strength to a residual value. This condition can be identified with
the presence of slickensided surfaces.

During the initial studies of Agua Vermelha Dam slickensided surfaces were recognized in the
first observations in pits, forcing to adapt the design to lower shear strength values. The volume
of this material involved was very large and its removal was not feasible. The chosen approach
was to perform a thorough study of the geotechnical properties and an intensive instrumentation
of the foundation (De Morais Leme et al, 1985). Aditionally, a stabilizing berm was built to
provide a factor of safety of 1 for residual strength. A secondary berm was foreseen in case
unacceptable movements occurred. After detailed observation, installed inclinometers
registered displacements as high as 25 mm, thus forcing the construction of the second berm to
avoid a progressive failure (De Morais Leme et al, 1985).

Fookes et al (1997) reported typical angles of shearing resistance for residual soils containing
residual clay minerals. The values are valid for soils in which the clay minerals are the
prevailing shearing mechanism (Fookes et al, 1997).

Clay Mineral φ’ CV φ’ R
Smectites 15-20° 5-11°
Kaolinites 22-30° 12-18°
Allophane 30-40° 30-40°
Halloysite 25-35° 25-35°

Table 6.4 - Shearing resistance angles according to clay mineralogy (Fookes et al, 1997).
 Page 37

6.4.5 Relic structures and discontinuities

As mentioned earlier the shear strength of saprolites resembles that of rocks, because of the
relic structures and discontinuities they inherit from the parent rock. These features are difficult
to identify by borings and drilling (Brenner et al, 1997). They are easier to detect through test
pits. Usually, saprolites fail under shear along these planes of weakness (Brenner et al, 1997).
Several authors have reported that the overall strength of the soils mass is controlled by the
prevalent “orientation and frequency of structural features in relation to the direction of stress
application, and to the strength characteristics of these features” (Brenner et al, 1997).

Relic structures are partly responsible for the anisotropic behavior of residual soils. Bedding
planes, schistosities, fractures and joints inherited from the parent rock are accountable for
different responses to shear, depending on the shear stress application in relation to the
structural features (Brenner et al, 1997). Wolle et al (1985) report shear stress parameters for
residual soils tested perpendicular and parallel to different types of structural features. Table 6.5
summarizes these results. Notice the large differences depending on the direction of testing.

Material Macro-structure Parallel Perpendicular Remarks

Saprolitic soil from Laminated c= 20 kPa c= 20 kPa Partially saturated
ferrifereous quartzite (Silty sand) φ=37° φ=37°
Saprolitic soil from Schistous c= 40 kPa c= 45 kPa Partially saturated
micaceous quartzite (Sandy silt) φ=22° φ=27°
Saprolitic soil from Banded c= 40 kPa c= 52 kPa Partially saturated
migmatitic geniss φ=20° φ=23°
Saprolitic soil from Banded c= 30 kPa c= 49 kPa Submerged
migmatitic geniss φ=21° φ=22°

Table 6.5 - Shear stress parameters for residual soils tested perpendicular and parallel to
structural features (Wolle et al, 1985).

Another consequence of the macro-structure of saprolites is that the stresses distribute in an

exceptionally heterogeneous manner within the soil mass. Great stresses concentrate on the
framework and remaining nucleuses (Wolle et al, 1985).

6.4.6 Partial Saturation

It was mentioned before that tropical regions where residual soils are commonly present often
present depressed groundwater tables (Brenner et al, 1997). Hence, tropical residual soils are
commonly not saturated. The pore pressures tend to be negative, due to capillary effects within
the pores of the soil (Brenner et al, 1997). As a result, the suction in the soil increases the
effective stress and consequently the shear strength. However, for design purposes this
additional component of strength should not be accounted, because it can easily disappear after
saturation.
 Page 38

6.4.7 Determination of shear strength parameters

Typical tests used for determining shear strength of transported soils can be utilized for residual
soils. The main differences lay not on the apparatus, but on the testing procedures and
interpretation. The main disadvantages with laboratory testing are the scale factor and
representativity. Commonly, the scale of relic structures exceeds that of the sample. Hence,
weak planes that frequently dominate shearing resistance in residual soils are not reflected in
the determined parameters. For laboratory tests the common direct shear tests and triaxial tests
are routinely carried out

Field testing has the benefit of reflecting the in-situ features that may dominate the shearing
behavior. Field direct shear test, Vane shear test, Pressuremeter test, c, and Cone Penetration
Test have been frequently used. Large shearbox tests are appropriate and give reasonable
results, but have the disadvantage of being costly and time consuming.

Back-calculating parameters from slope failures in the area can be a good approximation,
because a lot of aspects that difficult to take into account in test are incorporated in the analysis.

Correlations with index properties can be misleading and present large scatter. De Mello
(1985) warns against the errors and danger of using SPT (Standard Penetration Test) and dry
density indices to estimate foundation competence; particularly in soils exhibiting high
anisotropy (i.e. relic structures). For Itumbiara Dam (Brazil) SPTs were performed. Results
showed a large scatter (De Mello, 1985).

Typical values……To be continued

Guri Shearing strength Parameters

Antioquia Shearing strength Parameters
Brazilian Shearing strength Parameters
 Page 39

7 ADDITIONAL DESIGN CONSIDERATION

7.1 LATERITE AND SAPROLITE AS DAM FOUNDATION MATERIAL

7.1.1 TYPICAL PROBLEMS

7.1.2 FOUNDATION TREATMENT

7.1.3 INSTRUMENTATION

7.2 LATERITE AND SAPROLITE AS DAM FILLING MATERIAL

7.3 CONSTRUCTION METHODS

7.3.1 TIME RATE OF CONSTRUCTION

7.3.2 COMPACTION REQUIREMENTS

7.3.2.1 Layer Thickness

7.3.2.2 Lower vs. Higher than Optimum Water Content

7.3.2.3 Equipment

8 PERFORMANCE DURING OPERATION

8.1.1 PORE WATER PRESSURES

8.1.2 SEEPAGE

8.1.3 LONG TERM BEHAVIOR

8.1.4 STABILITY PROBLEMS

 Page 40

9 CONCLUSIONS

10 CASE HISTORIES

10.1 Colombian Dams

• Salvajina
• Piedras Blancas
• Quebradona
• Troneras
• Miraflores
• Santa Rita I
• Santa Rita II
• Punchiná
• San Lorenzo

10.2 Brazilian Dams

• Itá
• Machandinho
• Tres Marías
• Paraitinga
• Paraibuna
• Volta Grande
• São Simão
• Itumbiara
• Agua Vermelha
• Promisão
• Euclides Da Cunha Dam
• Graminha
• Limoeiro
• Jurumirim
• Barirí
• Jupía
• Chavantes
• Itaipú
 Page 41

10.3 Venezuelan Dams

• Guri

10.4 Additional Dams

• Babagon
• Ambuklao (Ask David Kleiner if the residual soils were removed)
• Binka
• Gangapur
• Gyobyu
• Samson Brook
• Sasumua

Please contribute specially sending additional case histories....

 Page 42

11 BIBLIOGRAPHY

De Mello V.F.B. & Mori R.T. “Dam foundations on tropical laterites and saprolites. General
Report”, International conference on tropical residual soils, saprolites and laterites, Brasilia,
Brazil 1985.

Fookes, Peter G editor “Tropical Residual Soils”, The Geological Society, London, 1997.

De Fries K., Prusza Z. & Tauseef C. “ Effect of Compaction on the Behavior of Residual
Soils” , Sixth Panamerican Conference ISSMFE, Lima, Perú, 1979

De Fries K., Prusza Z. & Tauseef C. “Field Compaction of Residual and Colluvial Gneiss
Derived Soils” , Sixth Panamerican Conference ISSMFE, Lima, Perú,1979

“Properties of Compacted Saprolites”, Sixth Panamerican Conference ISSMFE, Lima, Perú

Villegas, F “Remedial Works at Troneras and Miraflores Dams in Colombia”, Dam

Rehabilitation Practice

Vargas, M , Silva F.P. & Tubio M. “ Residual Clay Dams in the state of Sao Paulo, Brazil”,
Proc. 6th International conference on soil mechanics and foundation engineering, Montreal,
1966.

Nakao H. & Nahas C.M. “ Properties of Compacted Granite Saprolites”, Sixth Panamerican
Conference ISSMFE, Lima, Peru., 1979

Nogami, J.S. and others “Peculiarities of Tropical Lateritic and Saprolitic Soils used as
Construction Materials : Selection, Control and Acceptance Criteria.- Roads”, ISSMFE

De Moura, Leandro and others “ Mechanical and Hydraulic Properties of Tropical of

Tropical, Lateritic and Saprolitic Soils, Particularly as Related to their structure and Mineral
Components - Hydraulic Properties”

Lacerda, Willy and others “Mechanical and Hydraulic Properties of Tropical of Tropical,
Lateritic and Saprolitic Soils, Particularly as Related to their Structure and Mineral
Components - Compressibility Properties of Lateritic and Saprolitic Soils”, ISSMFE

Teixeira da Cruz, Paulo “Peculiarities of Tropical Lateriric and Saprolitic Soils used as
Construction Materials : Selection, Control and Acceptance Criteria - DAMS”, ISSMFE

Fernandes da Silva, Ricardo “Peculiarities of in situ behavior of tropical lateritic and

saprolitic soils in their natural conditions.- EROSION”, ISSMFE
 Page 43

De Morais Lemme, Clovis “Peculiarities of in situ behavior of tropical lateritic and saprolitic
soils in their natural conditions.- DAM FOUNDATIONS”, ISSMFE

Melfi Adolpho and others “ Characterization, identification and classification of tropical

lateritic and saprolitic soils for geotechnical purposes”, ISSMFE

Castro Alvaro, “Case history Saprolite as Foundation soil and Filling Material for Dams”, Not
Published.

Gigidasu, M.D., “LATERITE SOIL ENGINEERING”, Elsevier, Amsterdam,1976

Townsend, F.C. “Geotechnical characteristics of Residual Soils”, Journal of Geotechnical

Engineering, ASCE 111 (1): pp 77-94, January, 1985.

Wesley, L.D. “Influence of Structure and Composition on Residual Soils”, Journal of

Geotechnical Engineering, ASCE 116 (4): pp 589-603, April 1990.

Wesley, L.D. “Tjipanundjang Dam in West Java, Indonesia”, Journal of Geotechnical

Engineering, ASCE 100 (5): pp 503-522, May 1974.

Blight G.E. “Origin and formation of residual soils” in Mechanics of Residual soils G.E. Blight
Editor , A.A.Balkema; pp 1-15, 1997

Wesley L.D. and Irfan T.Y. “Classification of Residual Soils”, in Mechanics of Residual soils
G.E. Blight Editor , A.A.Balkema; pp 17-30, 1997

Queiroz de Carvalho J.B and Simmons J.V. “Mineralogy and Microstructure”, in Mechanics
of Residual soils G.E. Blight Editor , A.A.Balkema; pp 31-39, 1997

Simmons J.V. and G.E. Blight “Profile Description and Sampling Methods”, in Mechanics of
Residual soils G.E. Blight Editor , A.A.Balkema; pp 40-55, 1997.

Sowers George B, Sowers George F. “Introductory Soil Mechanics and Foundations”.

MacMillan Publishing Co., Inc. Third Edition, 1970.

Fourie A.B. , “Classification and Index Tests”, in Mechanics of Residual soils G.E. Blight
Editor , A.A.Balkema; pp 57-63, 1997

Simmons J.V. and G.E. Blight “Compaction”, in Mechanics of Residual soils G.E. Blight
Editor , A.A.Balkema; pp 65-77, 1997

Garga V.K and G.E. Blight “ Permeability”, in Mechanics of Residual soils G.E. Blight
Editor , A.A.Balkema; pp 79-93, 1997

Barksdale R.D. and Blight G.E. “Compressibility and Settlement of Residual Soils”, in
Mechanics of Residual soils G.E. Blight Editor , A.A.Balkema; pp 95-153, 1997
 Page 44

Brenner R.P., Garga V.K. and Blight G.E. “Shear Strength Behaviour and Measurement of
Shear Strength in Residual Soils”, in Mechanics of Residual soils G.E. Blight Editor ,
A.A.Balkema; pp 155-219, 1997

Blight G.E., “ Case Histories of Shear Strength - Controlled Aspects of Residual Soils”, in
Mechanics of Residual soils G.E. Blight Editor , A.A.Balkema; pp 65-77, 1997.

Townsend, F.C. 1985.”Geotehcnical characteristics of residual soils”. Journal of Geotechnical

Engineering, ASCE 111 (1): 77-94

ASCE (1971): Session 11: Slope Stability in Residual Soils, Proceedings, 4th Panamerican
Conference on Soil Mechanics and Foundation Engineering, San Juan, Puerto Rico, Vol. 2, pp,
105-167, and Vol. 3, pp. 125-146. Mitchell in-situ.

ASCE (1982): Proceedings of the ASCE Specialty Conference on Engineering and

Construction in Tropical and Residual Soils, Honolulu, 746 p.

AZZOUZ, A.S., BALIGH, M.M. & LADD, C.C.(1981): Three-dimensional stability analyses
of four embankment failures.
Mechanics and Foundation Engineering. Stockholm, Vol. 3, pp. 343-346.

BALKEMA (1981): Proceedings of the International Symposium on Laterisation Processes,

Trivandrum, India, 1979. Balkema, Rotterdam, 457 p.

BLIGHT, G.E. (1977): Slopes and excavations in residual soils.

Proceedings, 9th International Conference on Soil Mechanics and Foundation Engineering,
Tokyo, Vol. 3, pp. 582-590.

BRAND, E.W & PHILLIPSON, H.B. (1984): Site investigation and geotechnical engineering
practice in Hong Kong. Geotechnical Engineering, Vol. 15, pp. 97-153.

BRAND, E.W. &PHILLIPSON, H.B.(1985 a): Review of international practice for the
sampling and testing of residual soils.
Sampling and Testing of Residual Soils: A Review of International Practice, edited by E.W.
Brand & H.B. Phillipson, pp. 7-21. Scorpion Press, Hong Kong.

BRAND, E.W. & PHILLIPSON, H.B. (Editors) (1985 b): Sampling and Testing of Residual
Soils: A Review of International Practice. Scorpion Press, Hong Kong.

BRAND, E.W. & PHILLIPSON, H.B. BORRIE, G.W. & CLOVER, A.W. (1983 b): In-
situ direct shear tests on Hong Kong residual soils.
Proceeding, International Symposium on Soil and Rock Investigations by In-situ Testing, Paris,
Vol. 2, pp. 13-17. (Discussion, Vol. 3, pp. 55-56).

FREDLUND, D.G. &RAHARDJO, H. (1985): Theoretical context for understanding

unsaturated residual soil behaviour. Proceedings, lst. International Conference on
 Page 45

Geomechanics in Tropical Lateritic and Saprolitic Soils, Brasilia, Vol. 1, pp. 295-306.

GIDIGASU, M.D. (1975): Laterite Soil Engineering. Elsevier Scientific Publishing Company,
Amsterdam, 570 p.

HO, D.Y.F. & FREDLUND, D.G. (1982): Increase in strength due to suction for two Hong
Kong soils. Proceedings, ASCE Specialty Conference on Engineering and Construction in
Tropical and Residual Soils, Honolulu, pp. 263-295.

HUMA, I. & RADULESCU, D. (1978): Automatic production of the matic maps of slope
stability. Bulletin of the International Association of Engineering Geology, No. 17, pp. 95-99.

HUNT, T. (1982): Slope failure in colluvium overlying weak residual soils in Hong Kong.
Proceeding, ASCE Specialty Conference on Engineering and Construction in Tropical and
Residual Soils, Honolulu, pp. 443-462.

HUTCHINSON, J.N. (1877): Assessment of the effectiveness of corrective measures in

relation to geological conditions and type of slope movement. Bulletin, International
Association of Engineering Geology, No. 16, pp. 131-155.

KOO, Y.C. (1982 a): Relict joints in completely decomposed volcanics in Hong Kong.
Canadian Geotechnical Journal, Vol. 19, pp. 117-123.

KOO, Y.C.(1982 b): The mass strength of jointed residual soils.

Canadian Geotechnical Journal, Vol. 19, pp. 225-231.

LITTLE, A.L. (1967): Laterites. Proceedings, 3rd Asian Regional Conference on Soil
Mechanics and Foundation Engineering. Haifa, Vol. 2, pp. 61-71.

LITTLE, A.L. (1969): The engineering classification of residual tropical soils. Proceedings,
Specialty Session on Lateritic Soils, 7th International Conference on Soil Mechanics and
Foundation Engineering, Mexico City, Vol. 1 , pp. 1-10.

LUMB, P. (1962 a): The properties of decomposed granite. Geotechnique, Vol. 12, pp. 226-
243 (& 2 plates).

LUMB, P. (1965): The residual soils of Hong Kong. Geotechnique, Vol. 15, pp. 180-194.

MARTIN, R.P. & HENCHER, S.R. (1984): Principles for description and classification of
weathered rocks for engineering purposes. Proceedings, 20th Regional Conference of the
Engineering Geology Group of the Geological Society of London, Guilford, pp. 304-308.

MORGENSTERN, N.R. & DE MATOS, M. (1975): Stability of slopes in residual soils.

Proceedings, 5th Panamerican Conference on Soil Mechanics and Foundation Engineering,
Buenos Aires, Vol. 3, pp. 369-384.
 Page 46

MORI, H. (1985): Sampling and testing of granitic residual soils in Japan. Sampling and
Testing of Residual Soils: A Review of International Practice, edited by E.W. Brand & H.B.
Phillipson, pp. 99-108. Scorpion Press, Hong Kong.
OYAGI, N. (1984): Landslides in weathered rocks and residual soils in Japan and surrounding
areas. Proceedings, 4th International Symposium on Landslides, Toronto, Vol. 3, in press.

PATTON, F.D. & HENDRON, A.J. (1974): General report on “Mass Movements”.
Proceedings, 2nd International Congress on Engineering Geology, Sao Paulo, Vol. 2, pp. V-
GR.1 – V.GR.57.

PHILLIPSON, H.B. & BRAND, E.W. (1985): Sampling and testing of residual soils in Hong
Kong. Sampling and Testing of Residual Soils: A Review of International Practice, edited by
E.W. Brand & H.B. Phillipson, pp. 65-82. Scorpion Press, Hong Kong.

RICHARDS, B.G. (1985): Geotechnical aspects of residual soils in Australia. Sampling and
Testing of Residual Soils: A Review of International Practice, edited by E.W. Brand & H.B.
Phillipson, pp. 23-30. Scorpion Press, Hong Kong.

SOWERS, G.F. (1963): Engineering properties of residual soils derived from igneous and
metamorphic rocks. Proceedings, 2nd Panamerican Conference on Soil Mechanics and
Foundation Engineering, Brazil, Vol. 1, pp. 39-62.

SOWERS, G.F. (1971): Landslides in weathered volcanics in Puerto Rico. Proceedings, 4th
Panamerican Conference on Soil Mechanics and Foundation Engineering, San Juan, Puerto
Rico, Vol. 2, pp. 105-115.

SOWERS, G.F. (1985): Residual soils in the United States. Sampling and Testing of Residual
Soils: A Review of International Practice, edited by E.W. Brand & H.B. Phillipson, pp. 183-
191. Scorpion Press, Hong Kong.

TECHNICAL COMMITTEE ON TROPICAL SOILS (1985): Progress Report of the ISSMFE

Committee on Tropical Soils. ABMS, Brazilian Society for Soil Mechanics, Sao Paulo, in
press.

TERZAGHI, K. & PECK, R.B. (1948): Soil Mechanics in Engineering Practice. John Wiley
& Sons, New York, 584 p.

VARGAS M. (1953): Some engineering properties of residual clay soils occurring in southern
Brazil. Proceedings, 3rd International Conference on Soil Mechanics and Foundation
Engineering, Zurich, Vol. 1, pp. 67-71.

VARGAS M. (1967): Design and construction of large cuttings in residual soils. Proceedings,
3rd Panamerican Conference on Soil Mechanics and Foundation Engineering. Caracas, Vol. 2,
pp. 243-254.

VARGAS M..& PICHLER, E. (1957): Residual soil and rock slides in Santos (Brazil).
 Page 47

Proceeding, 4th International Conference on Soil Mechanics and Foundation Engineering,

London, Vol. 2, pp. 394-398.

VAUGHAN, P.R. (1985 a). Mechanical and hydraulic properties of in-situ residual soils.
Proceedings, lst International Conference on Geomechanics in Tropical Lateritic and Saprolitic
Soils, Brasilia, Vol. 3, 231-263.

VAUGHAN, P.R. (1988). Key note paper. Characterizing the mechanical properties of residual
soils. Proceedings, 2nd International Conference on Geomechanics in Tropical Soils,
Singapore 2: 469-487.
Page 48
 Page 49 from 49

Classification of Residual Soils (after

Wesley,1997)
Grouping system Common Descriptive information on in situ state
pedological names
used for group
Major Division Sub-group Parent Rock Information on structure
Group A (a) Strong macro- Give names if Give details of Describe nature of structure:
(Soils without a strong structure influence appropriate type of rock from -stratification, reflecting parent rock
mineralogical influence) which the soil -fractures, fissures, faults, etc.
has been derived. -presence of partially weathered rock
(state % and physical form, e.g. 50% corestones)
(b) Strong micro- Give names if . Describe nature of microstructure or evidence of it:
structure influence appropriate -effect of remolding, sensitivity
-liquidity index or similar index
© Little or no Give names if . Describe evidence for little or no structural
structure influence appropriate influence
Group B (a) Montmorillonite Black cotton soils, . Describe any structural effects, which may be
(Soils strongly influenced (Smectite group) Black clays, Tropical present, or other aspects relevant to engineering
by commonly occurring black earths, properties. Evidence of swelling behavior, extent of
materials) Grumusols, Vertisols surface cracking in dry weather, slickensides below
surface etc.
(b) Other minerals . . .
Group C (a) Allophane sub Volcanic ash soils, . Give basis for inclusion in this group. Describe any
(Soils strongly influenced group Andosols, Andepts structural influences, either macro-structure or
by clay minerals micro-structure.
essentially found only in
residual soils)
(b) Halloysite sub Tropical red clays, .
group Latosols, Oxisols,
Ferralsols
 Page 50 from 50

© Sesquioxide sub Lateritic soils, . Give basis for inclusion in this group. Describe
group (gibbsite, geo- Laterites, Ferralitic structural influences - Especially cementation
thite,haematite) soils, Duricrust effects of the sesquioxides.
 Page 51 from 51

Characteristics of residual soils groups (after Wesley,1997)

Group Examples Means of Comments on likely engineering properties and
identification behavior
Major Group Sub-group
Group A (a) Strong High weathered rocks Visual inspection This is a very large group of soils (including the
(Soils without a macro- structure from acidic or 'saprolites') where behaviors (especially in slopes) is
strong influence intermediate igneous dominated by the influence of discontinuities, fissures ,
mineralogical rocks, and sedimentary etc.
influence) rocks.

(b) Strong Completely weathered Visual inspection, These soils are essentially homogeneous and form a tidy
micro- structure rocks formed from and evaluation of group much more amenable to systematic evaluation and
influence igneous and sedimentary sensitivity, liquidity analysis than group (a) above. Identification of nature
rocks. index, etc. and role of bonding (from relict primary bonds to
weak secondary bonds) important to understanding
behavior.

© Little Soils formed from very Little or no This is a relatively minor sub-group. Likely to behave
structural homogeneous rocks sensitivity, uniform similarly to moderately overconsolidated soils.
influence appearance
Group B (a) Smectite Black cotton soils, Many Dark color (gray to These are normally problem soils found in flat or low
(Soils strongly (Montmorillonit soils formed in tropical black) and high lying areas, of low strength, high compressibility, and
influenced by e group) areas in poorly drained plasticity suggest high swelling and shrinkage characteristics.
commonly conditions soils of this group
occurring
minerals)
(b) Other This is likely to be a minor sub-group.
minerals
 Page 52 from 52

Group C (a) Allophane Soils weathered from Very high natural These are characterized by very high natural water
(Soils strongly group volcanic ash in the wet water contents, and contents, and high liquid and plastic limits. Engineering
influenced by tropics and in temperate irreversible changes properties are generally good, though in some cases high
clay minerals climates. on drying. sensitivity could make handling and compaction
essentially found difficult.
only in residual
soils)
(b) Halloysite Soils largely derived Reddish color, well- These are generally very fine-grained soils, of low to
group from older volcanic drained topography medium plasticity, but low activity. Engineering
rocks; especially tropical and volcanic parent properties are generally good. (Note that there is often
red clays rock are useful some overlap between allophane and halloysitic soils).
indicators.

© Sesquioxide This soils group loosely Granular, or nodular This is a very wide group, ranging from silty clay to
group re-ferred to as 'lateritic', appearence coarse sand and gravel. Behavior may range from low
or laterite plasticity to non- plastic gravel.