See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/259593142

Comparison between Stone Columns and Vertical Geodrains with Preloading

Embankment Techniques

Conference Paper · August 2008

CITATIONS READS

15 2,906

2 authors:

Mounir Bouassida Lassaad Hazzar

University of Tunis El Manar Hydro-Quebec
330 PUBLICATIONS   2,068 CITATIONS    29 PUBLICATIONS   307 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

calcul fiabiliste et stochastique des ouvrages View project

Techniques de télédétection, méthodes de traitement d’images et applications à l’étude des mouvements de terrain "SOUK AHRAS, ALGERIE". View project

All content following this page was uploaded by Mounir Bouassida on 06 November 2015.

The user has requested enhancement of the downloaded file.

 COMPARISON BETWEEN STONE COLUMNS AND VERTICAL GEODRAINS
WITH PRELOADING EMBANKMENT TECHNIQUES

Mounir Bouassida Lassaad Hazzar

URIG, ENIT URIG, ENIT,
Tunisia Tunisia

ABSTRACT

In the framework of “Radès-La Goulette“bridge project (Tunisia), this study focuses on the construction of embankments located in
north Lake of Tunis. These embankments with averaged height of about 6 m are founded on highly compressible clayey sand and
muddy sand layers. A soil improvement technique is then imposed, to overcome the lack of low bearing capacity and high pronounced
settlements. Two solutions of soil improvement have been studied; the first one consists in vertical “Geodrains” drilled until 10 m
depth associated with step by step construction of preloading embankment. The second technique is stone columns reinforcement up
to 10 m depth. It is focused at estimation of bearing capacity and prediction of settlement of reinforced soil by handling the recent
elaborated software programme “Columns”. The evolution of consolidation settlement of embankments as a function of time is also
considered. The consolidation of improved soil is studied by using the “poroelastic” prediction model and the Barron’s theory. A
comparison between the two soil improvement techniques from the technical and economical viewpoints is presented. Compared to
the “Geodrains” technique, the reinforcement by stone columns including the execution of embankments approximately leads to a gain
of eight months and slightly cost reduced.

INTRODUCTION The first technique (PVD) is well controlled and practiced in

Tunisia by the local entrepreneurs. In fact, PVD is a very
The study of embankments on compressible soils is one of the simple technique and it preserves the environment. Contrarily,
delicate problems which has been analysed by a large number the second technique (stone column reinforcement) should be
of authors. At the present time, in spite of all experience carried out by foreign entrepreneurs having long experience
obtained over the last decades, the stability of this kind of on the matter. This technique also requires advanced
constructions still collocates diverse and delicate issues as equipment for installation and to acquire stone material.
related to the weak bearing capacity, large settlements due to
high deformability and low permeability, and too slowly The geotechnical behaviour of an embankment on
dissipation of excess pore pressure (consolidation). compressible soils incorporating vertical drains or stone
columns is analysed during and after the construction period.
Designing embankments on highly compressible soils usually The first part of this paper is dedicated to analysis of
involves soil improvement techniques as useful alternative geotechnical data which includes the classification and
which allow a reasonable duration of construction especially interpretation of laboratory and in situ tests results.
for big projects in coastal areas which basically includes land
reclamation. The second part focuses on the consolidation of highly
compressible layers as foundation of high embankments.
The big project “Radès-La-Goulette” bridge which connects Prefabricated vertical drains associated with a preloading
the north and south parts of the capital Tunis (fig. 1) surcharge are then undertaken.
comprises four lots. Part of them is the construction of four
embankments of access in north lake area. In order to ensure In the third part, a stone columns reinforcement of
the stability of embankments which final height varies from 5 compressible layers is suggested and followed by appropriate
to 6.5 m, two soil improvement techniques have been studied: design by using the software programme “Columns”.
prefabricated vertical drains (PVD) with preloading surcharge
and stone column reinforcement. Finally, a technical-economical comparison between the two
soil improvement alternatives is presented, which makes it

Paper No. 7.18a 1

possible to decide the adequate solution to build the Only the results obtained from borehole (ard1), located in the
embankments of access. area of North Lake are exploited (fig. 2).

In this paper, the French abbreviation NGT means “General The first geotechnical synthesis displayed a very soft
levelling of Tunisia”. compressible layer I of thickness varying from 8 to 10 m.

Fig.1. Location of the project “Radès-La Goulette” bridge.

GEOTECHNICAL INVESTIGATION OF THE SITE

“Radès-La Goulette “bridge is a big Tunisian project which

Fig.2. Results of CPT (qc in MPa) vs depth in meter.
comprises several parts. A part of them is the construction of
embankments with variable height up to +8 m NGT located
This soft layer can be divided into two sublayers: the first is a
behind of the abutments of bridges of access; in the north
highly compressible mud of about 5 to 6.5 m thickness, while
lake’s zone currently presenting a draught of about 1m (the
the second sublayer consists in compressible sandy clays.
depth of Bed Lake is located between -1 to -0.6 m NGT).
The recorded data static cone test from penetration confirmed
the existence sublayer Ib.
The soil profile, as foundation of embankments, presents
successive clayey and sandy deposited under consolidated
It also noticed a moderate difference between the cone
layers. However, the continuity in horizontal plan of these
penetration resistance and undrained cohesion in the two
alternations, along 22 to 25 m depth, is not necessarily proved.
sublayers Ia and Ib.
Distances between bored holes ranges between 100 m to 200
m.
For studying the stability and preloading process, a focus on
the behaviour of levels I and III has been addressed (fig. 2).
Geotechnical Model
Table 1 groups the significant data related to mechanical
characteristics of the soil as foundation of the studied
The geotechnical model is set up based on results obtained
embankments.
from: bored holes, and in situ tests (pressuremeter and SPT
data) including the static cone penetration records and pore
water pressure (piezocône). The first geotechnical
investigation (boreholes, drilled core samples, pressuremeter
tests, vane tests, SPT) was mostly conducted along various
depths (40 m to 110 m) (NIPPON KOE et al, 2001).

Paper No. 7.18a 2

Table 1. Mechanical characteristics from in situ tests. reinforcement alternative has the advantages of significant
reduction of long-term settlement and the construction of
Location Layer Tip resistance Undrained embankment of access will enhance significant increase of
qc (MPa) cohesion bearing capacity due to mechanical performances of columns
Cu (kPa) material.
Reference of Ia 0.025+0.032 z 1.5+2.1 z
boreholes Ib 0.2 13
III, VI Refusal at 10 m STUDY OF THE EMBANKMENTS OF ACCESS
Reference of in Ia 0.19 13
situ tests Ib 0.25 17 The main difficulties which arise for the construction of
III, VI 11.4 z 0.76 z embankments of access are:
- Short-term stability of the soft ground as related to bearing
Results capacity verification.
- Long-term settlement of unimproved deep layers (depth
The sand layer located at 12 m depth has unsignificant greater than 30 m).
plasticity, which indicates its negligible compressibility.
Description
The characteristics of compressibility, measured from
oedometric test, for layer I (until 10 m depth) are presented in The zone of the interchange which comprises the new express
Table 2. route and four embankments of access, cover approximately
16 hectares to be reclaimed in the north lake of Tunis. The
Table 2. Adopted characteristics of compressibility for soft final heights (after end of primary consolidation) of these
soil. embankments vary from 5 m to 6.5 m.

Layer Elevation γ (kN/m3) Cc/1+ e0 Cv (E-08 tg φcu Based on predicted settlements, under centre line of each
(m) m²/s) embankment of access, by the odeometric and pressurmeter
Ia -0.9 – 6.5 16.5 0.15 5 0.158 methods, the height of preloading embankments was deduced.
Ib -6.5 – 9.2 19 0.1 8.8 0.158 Because of too low short-term mechanical characteristics of
II -9.2 – 18.8 18.5 0.09 - -
III -18.8 –26.8 18 0.14 2 to 4 -
the foundation of embankments, a staged construction is
IV -26.8 - 35 19 0.05 - - scheduled. Such a procedure will make possible the increase
V -35 - 71 18.8 0.18 5 - of short-term shear strength of soft layers as consequence of
part of the primary consolidation.
The horizontal coefficient of compressibility Ch is estimated
from the vertical coefficient Cv as: Ch = 5Cv. As potential soil improvement techniques achievable in the
context of "Radès La Goulette" bridge project, the design will
Due to the significant lack of bearing capacity and the high be proceeded, first, for the prefabricated vertical drains (PVD)
compressibility of soil layers along 10 to 15 m depth, the associated to preloading embankments and, second, for the
construction of embankments is definitely compromised. stone columns reinforcement.
Besides, significant settlements are also predicted in
compressible deep layers (levels III & IV). For these reasons Stability of embankments
making recourse to an improvement solution of soils layers
under the embankments, at least along the first 10 m depth, The slopes of embankments of access are projected as 3 m for
reveals unavoidable. horizontal and 1 m for elevation. The platform is located at + 1
m above the NGT level.
Such a solution aims, first, the acceleration of consolidation of
high compressible layers. In case a reinforcement technique The fill material used has an angle of internal friction of 30°,
might be envisaged a significant reduction of settlement consequently tg 30° = 0.57 > 1/3. Then, a safe stability of
associated with substantial increase of bearing capacity will be slope embankments is guaranteed. The in situ unit weight
possible. Then, the two alternatives soil improvement embankment’s compacted material is about of 19 kN/m3.
techniques are:
- The use of vertical geodrains associated with preloading Staged construction of embankments
embankment.
- The soil reinforcement by stone columns (or by sand piles). The stages of construction of embankments have been
scheduled as follows:
Each of the two alternatives has specific advantages. Indeed, • Reclamation of the total area by a generalized fill at
by the technique of geodrains, which is characterized by a +1 m NGT.
rapid installation, the consolidation of soft ground is well • Arise the thickness of embankments of access, at + 3
accelerated. Meanwhile, a staged construction for m to + 5 m NGT: in zone of connection with the
embankments is necessary. Whereas the stone columns express route.

Paper No. 7.18a 3

 • The thickness of embankments of access behind the Preloading
abutments, along 20 m length, is arisen at +8.0 m
NGT. Preloading refers to the process of edification of a temporary
• embankment prior to the placement of final permanent
The allowable bearing capacity complying with the initial construction. If the temporary applied load exceeds the final
height of the embankment to build is: Hr = 2m. In turn, the loading, the amount in excess is referred to be as surcharge
construction of an embankment with height exceeding 2m load. Since the preloading is rapidly applied, the resulting
requires a soil improvement solution. settlement of soft mud deposit is divided into immediate and
primary consolidation components. This latter generally
The consolidation of the sandy mud layer by the technique of predominates because of the negligible immediate settlement
preloading revealed insufficient. Indeed, for the 2 m initial compared to that of primary consolidation.
height of preloading, the time of primary consolidation of the
mud layer is about of 58 years, which is quite inadequate with The preloading steps are designed based on the gain of
the duration of embankments construction. undrained cohesion which results from the accelerated
consolidation of high compressible layers. The increase of
A first adequate solution consists in associating with the initial undrained cohesion ∆Cu , due to a prefixed degree of
preloading a network of prefabricated vertical drains in order consolidation U (%) , will serve to design the next preloading
to accelerate the consolidation of the mud layer. However, step; then:
using the stone columns reinforcement technique, an increase
of the bearing capacity, and significant settlement reduction of ∆Cu
reinforced soil will be provided, adding to the acceleration of = U (%) (1)
consolidation enhanced by the drained nature of columns γ H tgΦ CU
material.
tgΦ CU = the rate of increase of undrained cohesion under the
effect of the consolidation. After available data related to the
IMPROVEMENT BY PVD WITH PRELOADING soft ground of Tunis, we have tgΦ CU = 0.158 (Bouassida,
EMBANKMENT
2006).
The prefabricated vertical drain (PVD) with preloading
method was considered as the most feasible treatment for the Table 3 presents the increase of undrained cohesion
project based on the depth of treatment, cost, allocated time occasioned in the sandy mud as a result of the primary
for preloading and other considerations (fig. 3). The objective consolidation which occurs at the end of each preloading step.
of using vertical geodrains with preloading technique is to
accelerate the rate of consolidation and to minimize the Table 3. Increase of undrained cohesion of sandy mud layer
remaining settlement of the treated area under the final (dead resulting from staged preloading.
and live) loadings. Preloading increases bearing capacity and
reduces the compressibility of weak ground by forcing soft Elevation (m) Total height of ∆CU (kPa ) U r (%)
soils to consolidate (Van Impe, 1989). Soil improvement NGT embankment
works is carried out in such a way that a specified degree of H(m)
primary consolidation is designed to be attained during the -1 0 0 0
desired time by improving the soil drainage system. 0.5 1.5 0 0
1 2 0 0
3 4 0 0
5 4.5 5.5 53
5 5 6.5 54
5 5.5 6 45
5 6 6 40
8 9 11 67

The primary consolidation settlement in centre line of

embankment of sandy mud layer assumed as normally
consolidated is predicted after one dimension Terzaghi’s
theory:

Fig.3. Construction sequences of preloading embankment

on improved soil by PVD.

Paper No. 7.18a 4

 ⎛ C H ⎞ ⎛ σ ' + ∆σ ⎞ Choice of the drains pattern and preloading schedule
s∞ = ⎜ c 0 ⎟ log ⎜ 0 ' v ⎟ (2)
⎝ 1 + e0 ⎠ ⎝ σ 0 ⎠ The waiting time between successive stages of preloading has
been determined for a given degree of horizontal consolidation
∆σ v = excess of vertical stress. calculated by Barron’s formula.
1
C c = index of compression. ln( )
⎡ D4 D 3D 2 − d 2 ⎤ 1−Uh (5)
t=⎢ ln( )−
σ 0' = effective overburden stress at night of compressible layer ⎣ 8( D − d ) d
2 2
32 ⎦⎥ Cr
Consider data: H0 = 6.5 m; ∆σv = 2×19 = 38 kN/m²; then
s∞ = 1.0 m t: time in seconds;
Uh: degree of horizontal consolidation in %;
The degree of consolidation of foundation layers beneath the
D = 1.13 L; D: equivalent diameter, L: spacing core to core
embankments of access is approximated by:
between drains installed in square pattern.
st The waiting time between preloading stages varies from 35 to
U (t ) = (3)
s∞ 70 days for a squared pattern drains spacing of 1.8 m. This
corresponds to the agenda planned of the site reclamation,
s∞ and st denote respectively the settlements at the end of without making recourse to an accelerated consolidation with
primary consolidation, and at given time, which corresponds a tighter platform. The total duration to attain the level + 8 m
to the degree of primary consolidation U (t ) . NGT is 245 days.

Characteristics of PVD However, the fact of adopting a tight grid of 1.2 m spacing,
under the most loaded zones, with a transition zone with a grid
A prefabricated vertical drain (PVD) can be defined as any of 1.5 m spacing, makes it possible to anticipate settlements
prefabricated material or product consisting of synthetic filter behind the abutments of bridges access. The total waiting time
jacket surrounding a plastic core having the following is 89 days, which corresponds to 63% of the time expected for
characteristics (Bergado et al, 1996): a pattern where 1.8 m spacing is adopted.
- Ability to permit porewater in the soil to seep into the
geodrain. Meanwhile, in the two cases, the elevation of embankment
- A tool by which the collected porewater can be transmitted +3m NGT level does not require significant waiting time (less
along the length of the geodrain, without any particles than 15 days).
migration from the soil to improve during drainage.
Tables 4 and 5, and curves illustrated in figure 4 give the
The studied case history considers an acceleration of the predicted aimed waiting time to acquire the improvement of
consolidation by the installation of a grid of PVD descended the north Tunis lake area to be reclaimed.
from a platform levelled +0.5 m NGT, until 10 m depth. The
proposed type of PVD is Mebradrain 88 (MD 88) which is of Table 4. Drains installation with spacing 1.8 m.
flat type of thickness 0.5 cm and 10 cm width. MD 88 was
also experienced in previous soil improvement projects with Elevation (m) Hr (m) Waiting time Cumulated
PVD in Tunisia (reclamation in South Lake of Tunis). /NGT (days) time (days)
From -1 to +0.5 1.5 0 0
A 0.5 m thickness drainage blanket made up of gravel-sand From +0.5 to 2 0 0
material will cover the improved soft layer to speed the PVD +1
drained water and will serve as platform for settlement From +1 to +3 4 0 0
recorders, piezometers. From +3 to 4.5 50 50
+3.5
The geometrical and hydraulical characteristics of PVD are: From +3.5 to 5 50 100
- A diameter of the drain: +4
From +4 to 5.5 40 140
perimeter 2 × (10 + 0.5) +4.5
d= = = 6.7cm (4)
π π From +4.5 to 6 35 175
+5
- A capacity of discharge: From +5 to +8 9 70 245
q w = 5.10−5 m3 s −1 .
- A mass: 96 g/linear meter.

Paper No. 7.18a 5

Table 5. Drains installation of 1.2 m spacing.

Elevation (m) Hr (m) Waiting time Cumulated 120,0

/NGT (days) time (days) 100,0

From -1 to 1.5 0 0
+0.5 80,0

From +0.5 to 2 0 0

s (cm)
60,0 1.2 m
+1 1.8 m
From +1 to +3 4 0 0 40,0

From +3 to 4.5 18 18 20,0

+3.5
0,0
From +3.5 to 5 19 36 0 50 100 150 200 250 300 350 400 450
+4 Time (days)
From +4 to 5.5 14 50
+4.5
From +4.5 to 6 12 63
Fig. 5. Evolution of settlements and influence of spacing vs to
+5
time of preloading.
From +5 to +8 9 26 89
It is also worth mentioned to predict the gain of the undrained
cohesion of soft layer after improvement by PVD installation.
Evolution of settlements versus time of preloading
Such a result will serve for studying the stability of the
foundation’s layer of embankment.
In order to determine the evolution of settlements during the
whole staged construction of embankments of access, the
variation in time of the degree of consolidation is studied for REINFORCEMENT BY STONE COLUMNS
each soil level (layer).
The stone column technique was adopted especially in
0 30 60 90 120 150 180 210 240 270 300
European countries early in the sixties and became little by
9 little successfully practiced. A stone column is basically a
8 vertical cylindrical “hole” executed in a soft soil layer and
7 filled with compacted stone fragments and gravel having high
6 potential drainage.
Elevation m NGT

5
4 This technique can be used to improve soft layers under dams
3 and embankments in order to increase the bearing capacity, to
2
reduce settlements, and to accelerate the consolidation process
1
1.8 m 1.2 m like vertical drains.
0

-1
-2
Stone columns are basically installed either by the use of vibro
replacement or by use the vibro displacement process. Figure
Tim e (days)
6 depicts the different stages of a process stone column
installation by, the vibro displacement. More detailed
descriptions of the equipment and the procedure itself can be
Fig. 4. Stages of embankment’s construction vs time. found in Moseley & Priebe (1993), Kirsch & Sondermann
(2003), Debats (2006).
Figure 5 presents the evolution of settlements of primary
consolidation of layers Ia and Ib by taking into account the
effect of accelerated consolidation which results from a
pattern of squared PVD when the spacing takes 1.2 m and 1.8
m. In fact, figure 5 shows up effectiveness of reduced spacing
in the gain of time of consolidation to reach the same
magnitude of settlement.

Paper No. 7.18a 6

 angle of internal friction of the soil is ϕ = 16°, then the
minimum improvement area ratio is ηmin = 16.7 %.
2. The prediction of ultimate bearing capacity (qult) refers to
the case of purely cohesive soils reinforced by cohesive
frictional columns material. Two methods of design, are
involved, namely, the yield design approach (lower
bound) which takes account of improvement area ratio
and the recommendations of French Standard “NFP 11-
212, (2004)” which do not take account, of improvement
area ratio. Then the allowable bearing capacity is deduced
based on a given global safety factor which depends on
the method of prediction (Table 7).

Table 7. Comparing between predictions of allowable bearing

Fig.6. Dry bottom feed vibro displacement method.
capacities.
Usually the columns are placed in a regular pattern, squared or
triangular, improving the weak layers below the embankment. Methods Global safety qall [kPa]
factor
As basic parameter for the design of column reinforcement is Yield design 1 114
the improvement area ratio defined by: (lower bound)
NFP 11 -212 2 177
Ac (6)
η= 3. The verification of the allowable bearing capacity with
A
respect to the embankment load led to a minimum
A : Area of foundation. improvement area ratio: η≥16.7 %. The evolution of
Ac : Total cross section of columns located under the loading lower bound ultimate bearing capacity as a function of
improvement area ratio is illustrated in Figure. 7.
foundation.

Properties of column material

Stone columns are usually installed using deep vibratory

compaction equipment (vibro probe). Columns material is
generally acquired from a quarry, i.e. selected crushed gravel
having a prescribed grain size. Drained characteristics of stone
columns material (Costet and Sanglérat, 1983) are grouped in
Table 6.

Table 6. Characteristics of stone columns material.

γ [kN/m3] C’ [kPa] E’ [MPa] ϕ’ [°] ν’

20 0 10 40 0.25

For the present case history, stone columns are designed with
final diameter of 1m to be installed pre-bored holes along 10 Fig.7. Evolution of the ultimate bearing capacity versus
m depth using a vibro displacement method (fig. 6). improvement area ratio (output of software “Columns”).

Bearing capacity
Settlement predictions
The bearing capacity of a supported foundation is the vertical
stress which causes the yield of underlying soil of foundation. Presently, available methods for settlement prediction can be
classified either as simple methods which use the one
For embankments of access, the bearing capacity verification dimension linear elastic model assumptions or as sophisticated
has been designed by using the too recent elaborated software methods using numerical codes which consider linear elastic
“Columns” (Bouassida et al, 2007) as detailed below. and/or elasto-plastic behaviour 2D or 3D model.
For this project, the prediction of settlement is carried out by
1. The minimum improved area ratio ηmin is predicted based using the software “columns” (Bouassida et al, 2007), in
on the limit analysis approach (Bouassida, 2007). The which the linear elastic behaviour is adopted by several
methods of design for constituents of reinforced soil.

Paper No. 7.18a 7

 Ea and Es denote respectively the apparent modulus of
Settlement before reinforcement is about 2.1 m in the center reinforced soil and Young modulus of initial soil.
line of embankment having 6 m height. Meanwhile, the
admissible settlement should not exceed 30 cm. It is observed a quasi linear relationship for all methods of
prediction.
The settlement, in center line of embankment, generated by
the load of final height of embankment (Hr = 6 m), is
estimated, assuming the linear elastic behaviour, by the
variational approach and French recommendations NFP 11-
212 (Table 8).

For each method of design, the settlement complying with

admissible bearing capacity is estimated, first, by considering
the minimum improvement area ratio and, second, by the
optimized improvement area ratio which complies with
allowable settlement.

Table. 8. Comparison between predicted settlements by two

methods.

Methods Variational approach NFP 11-212

Fig. 9. Variation of Normalized Young modulus of reinforced
η = 16.7 %. 45.5 cm 39.5 cm
soil versus improvement area ratio (η).
η = 31.5 %. 30 cm 28 cm

According to the height of embankment, or conversely the Design of stone columns network
applied load, it is possible with “columns” software to predict
the variation of settlement by several methods all assuming The stone columns network has been designed with specific
columns of end-bearing type (fig .8). The most conservative parameters grouped in table 9.
prediction is given by Chow’s method which uses the unit cell
model and assumes zero horizontal displacement in each point Table. 9. Designed stone columns network.
of soil reinforced. While the variational method uses the
group of columns model and takes account of lateral Length Substitution Columns Spacing Pattern
confinement in 3D reinforced soil (Bouassida et al, 2003). (m) factor (%) diameter (m)
(m)
10 31.5 1.0 1.7 Triangular

Consolidation

Stone columns also behave as vertical drains and because of

the drained property of their constitutive material which
accelerates the process of consolidation.
In order to predict the evolution of settlement versus time,
performing the poroelastic approach (Guetif and Bouassida,
2005) programmed in “columns” software, the evolution of
settlement of columnar reinforced soil is predicted as a
function of the history of loading.
Horizontal permeability is the needed parameter for carrying
the poroelastic approach:
Fig.8. Settlement of reinforced soil versus applied load.
Ch ⋅ γ w (7)
The apparent normalized Young modulus of the reinforced kh =
ground is represented in Figure 9 with other modulus Eoed
estimated by linear elastic method: variational method
(Bouassida et al, 2003), (Balaam & Booker, 1981), (Chow,
1996) and (NFP 11-212, 2004), as a function of the
improvement area ratio.

Paper No. 7.18a 8

Odeometric modulus is: COMPARISON BETWEEN SOIL IMPROVEMENT
TECHNIQUES
(1 −ν s )
: E oed = Es (8) In this study, for materials to be acquired and installation
(1 − 2ν s )(1 +ν s ) techniques, the current costs in Tunisia are applied.
ν s : Poisson’s ratio.
Soil improvement by PVD associated with embankment of
γ w : Unit weight of water. preloading
NA: k h = 1.67 10 −9 m / s .
Prediction by the poroelastic approach illustrated in Figure 10 The construction of embankments of access is carried out by
shows the evolution of settlement versus time, and indicates using a selected light weight material (expanded clay of unit
that the final consolidation settlement is expected in 97 days. weight = 7 kN/m3) along 10 m length behind the abutments,
and the volume of filled material is of about 2500 m3. Table
10 summaries the improvement technique by PVD associated
with preloading embankment. Note, that the time of execution
is about sixteen months.

Table 10. Cost of execution of soil improvement by PVD.

Volume of embankment of preloading 48310 m3

Volume of material to acquire 17843 m3
Volume of weight light material 2467 m3
Linear meter of PVD 57810 lm.
Cost (TND) 1,472,090

Stone column reinforcement technique

The predicted settlement at end of construction of

embankments on columnar reinforced soil is about of 40 cm.
Fig. 10. Settlement evolution of reinforced ground versus time. Then, consumption added column material is required for the
definitive embankments and consequent cost follows. Table
The evolution of consolidation settlement is greatly influenced 11 indicates the cost of execution of the reinforcement
by the value of substitution factor as shown in Figure 11. technique by stone columns.
Meanwhile for a wide margin of the substitution factor,
currently practiced for stone columns technique, the end of Table 11. Cost of soil reinforcement by stone columns.
primary consolidation of reinforced soil in average takes 150
days. Note that for low values of improvement area ratio (less Volume of embankment of access 17493 m3
than 10%) predicted settlement by the poroelastic approach is Linear meter of columns 3690 m
not realistic. Cost (TND) 1,226,955

Economical comparison between the two alternatives

The estimated cost for the alternative “Soil improvement by

PVD with preloading embankment” is about 1,472,090 TND
which includes the cost of preloading and unloading
embankments, and the cost of geodrains. The cost of
installation of linear meter of MD 88 geodrains is 2.5 TND.
The estimate of the second alternative “Reinforcement by
stone columns” technique is about 1,226,955 TND. The cost
of installation of a stone column linear meter is by 80 TND.

From Tables 10 & 11, it is clear that the column reinforcement

technique provides a reduction of about 16.6% on the cost of
the foundation under embankments of access. The time of
Fig. 11. Variations of the settlement of reinforced soil vs time, execution is of about eight months which provides a
for variousη. substantial gain of eight months compared to PVD installation.
Because PVD improvement includes preloading and
unloading steps during embankments it takes a longer time of

Paper No. 7.18a 9

construction than that estimated for stone columns
reinforcement.

Multi criteria analysis

Table12 recapitulates the specifications of each improvement

technique envisaged during the execution of project "Radès La
Goulette Bridge" for the foundations of the exchanger in north
Lake of Tunis.

Table 12. Multi criteria analysis of studied improvement

techniques.

Techniques Improvement with Stone column Fig. 12. Preparing PVD installation.
PVD reinforcement
Qualification of Very good less
local entrepreneurs
Duration of Long (16 months) Short (8 months)
execution
Environnemental unsignificant unsignificant
impact
Cost normal Less important
Comments Well controlled Little use

The multi criteria analysis highlights the economical interest

(cost and time of execution) shown by the stone column
technique which appears more advantageous than PVD with
preloading embankment.

Despite the advantages in favor of stone columns

reinforcement technique, the owner of “Radès La Goulette”
Fig. 13. Starting PVD installation.
bridge project decided the execution of PVD as improvement
solution. Such a choice is justified based on a much better
qualification of Tunisian entrepreneurs for PVD installation
and, in parallel, few practice of stone columns installation.

GEOTECHNICAL SURVEY

The geotechnical survey in reclaimed north lake area has been

instrumented by installed piezometers and settlements plate
readings located along the cross section of embankments of
access.

Figures 12, 13, 14, 15, 16 and 17 illustrate successive

operations of PVD’s installation and location of in situ record
instruments.

Fig. 14. PVD fixed to mandrel.

Paper No. 7.18a 10

 Fig. 15. Installed settlement plate.

Fig. 17. Locations of settlement recorders in PVD

improved soil.

Recorded in situ measurements are still continuing (until end

of 2007). The first results inform consolidation settlement is
not completely stabilized under embankments of access which
end of construction was in March 2007. For this comparison
between predicted evolution and observed settlements did not
yet start.

CONCLUSION

Because of the mediocre characteristics of the Tunis subsoil

along the first twenty meters, a soil improvement solution has
been decided to make possible the construction of
embankments of access in the north lake area of Tunis as part
of the big project “Radès-La Goulette” bridge.

Two solutions of soil improvement have been studied; the first

one consists in vertical “Geodrains” drilled until 10 m depth
Fig. 16. Protected settlement recorder. associated with step by step construction of preloading
embankment, the second solution is stone column
reinforcement up to 10 m depth.

• Improvement by PVD:
It revealed, when associated with preloading embankments, as
convenient solution to reach a high degree of primary
consolidation. Consequently, major part of settlement will be
released during the period for construction of embankments of
access. Effectiveness of PVD soil improvement largely
depends of adopted spacing between drains.

• Stone columns Reinforcement:

The gain in time of execution and subsequent economical cost
are in favour of this reinforcement technique which guarants

Paper No. 7.18a 11

significant increase of bearing capacity, decrease in Costet J. and Sanglérat G., [1983]. “Cours pratique de
consolidation settlement, adding to accelerated consolidation mécanique des sols, Calcul des ouvrages”, Vol.2, Edit. Dunod,
Paris.
A multicreteria analysis comparing between the two
improvements techniques highlighted the stone columns as Chow Y.K., [1996]. “Settlement analysis of sand compaction
more advantageous essentially the economical viewpoint. In Pile”, Soils and Foundations, Volume 36 - n°1, Japanese
turn, the PVD technique, being more experienced in similar Geotechnical Society, 111-113.
previous project of reclamation by Tunisian entrepreneurs,
was finally decided for execution. Guetif Z. and Bouassida M., [2005]. “Analytical estimate of
settlement evolution of soft soil reinforced by stone columns”,
Authors gratefully acknowledge the Tunisian Ministry of Proc. 16th International Conference of Soil Mechanics and
Equipments “de l’habitation et de l’aménagement des Geotechnical Engineering. September 12th-16th, Osaka, Japan,
territories”, and “Hydrosol-Foundations S.A” (Tunisia) for Vol. 3, pp 1355-1358.
provides useful data geotechnical and for providing helpful
data as related to geotechnical investigations and to PVD Hydrosol [2003]. Geotechnical Investigations in the Tunis
installation. Lake – North zone C. Project Radès-La Goulette bridge,
Tunisia.

REFERENCES Indraratna B., Sathananthan I., Rujikiatkamjoorn C., and

Balasurbramaniam A.S., [2005]. “Analytical and numerical
Barron A. [1947]. “Consolidation of fine grained soils by modelling of soft soil stabilised by PVD incorporating vacuum
drains wells”, American Society of civil Engineers, Journal of preloading”, International Journal of Geomechanics, Volume
Soils Mechanics, Vol.73, SM6, 718-743. 5, n°2, 114-124.

Balaam N.P., Booker J.R., [1981]. “Analysis of rigid rafts Kirsch, K. and Sondermann W., [2003]. “Ground
supported by granular piles », International Journal for improvement”, In U. Smoltczyk (ed.), Geotechnical
Numerical and Analytical Methods in Geomechanics”. Engineering Handbook. Vol.2, 1-56, Berlin: Ernst & Sohn.
Volume5- n°4, 379-403.
Moseley M.P., Priebe, H.J., [1993]. “Vibro techniques”, In
Bergado D.T., Anderson L.R., Muira N., Balasubramaniam M.P. Moseley (ed.), Ground Improvement: 1-19.Glasgow:
A.S., [1996]. “Soft Ground Improvement in Lowland and Blackie.
other environments”, Edition ASCE, New York.
Madhira R., Madhav and Nagpure D. D. [1996]. “Design of
Bouassida M., Guetif Z., De Buhan P., Dormieux L., [2003]. Granular Piles for Embankments on Soft Ground”, 12th South
Estimation par une approche variationnelle du tassement d'un East Asian Geotechnical Conference, Kuala Lumpur.
sol renforcé par colonnes. Revue Française de Géotechnique. NIPPON KOE, STUDI and SCET-Tunisie, [2001].
n ° 102, pp 21-29. “Document of geotechnical investigation”, detailed preproject
for “Radès- La Goulette” bridge constructed.
Bouassida M., [2006]. “Modelling the behaviour of soft clays
and new contributions for soil improvement solutions”, Proc. Normes Françaises NFP 11-212, [2004]. “Recommendations
2nd Int. Conf. on Problematic Soils, December 3rd-5th 2006. on the design, the calculation, the execution and the control of
the stone columns under buildings and sensitive structure”.
Bouassida M., [2007]. “A novel design methodology for COPREC and SOFFONS. Paris.
column reinforced soil”, Power point presentation in TC17
Workshop. Madrid, September 24th, 16th European Conf. Soil Poorooshasb, H.B. and Meyerhof, G.G., [1993].
Mech. & Geotechnical. Eng. “Consolidation Settlement of Rafts Supported by Stone
Columns”, Ground Engineering Journal, 47-70.
Bouassida M., Hazzar L., De Buhan P., [2007]. “Software for
designing foundations on reinforced soil by columns” (in Priebe H. J., [1995]. “The Design of Vibro Replacement”.
French), Proc.14th African Regional Conference Yaoundé Ground Engineering Journal.
2007. (In print)
Priebe H. J., [1998]. “Vibro Replacement to Prevent
Debats J.M., [2006].”Exécution et suivi de projets de Earthquake Induced Liquifaction”, Geotechnique-Colloquium,
renforcement par colonnes ballastées ; colonnes en sol traité Darmstadt, Germany.
aux liants”. Invited lecture in « Ground improvement »
Touring lecture. Hammamet, December 15-16th, Tunisia. Van Impe W.F., [1989]. “Improvement Techniques and their
Evolution”, Balkama, pp. 125

Paper No. 7.18a 12

View publication stats