Sei sulla pagina 1di 160

Volume Prepared by ISSMGE Technical Committee - 214

Volumen preparado por el Comit Tcnico TC-214 de la ISSMGE

For / para

3rd International Conference on Deep Foundations


Deep Foundations and Soil Improvement in Soft Soils

3er Simposio Internacional de Cimentaciones Profundas


Cimentaciones Profundas y Mejoramiento Masivo en Suelos Blandos

November 11-12th, 2015, Mexico City

Edited by / Editado por


Norma Patricia Lpez Acosta

Technical Committee

TC-214

Copyright, Mxico, 2015


Sociedad Mexicana de Ingeniera Geotcnica, A.C.
Valle de Bravo No. 19 Col. Vergel de Coyoacan,
14340 Mxico, D.F., MXICO
Tel. +(52)(55)5677-37-30, Fax+(52)(55)5679-36-76
Pgina web: www.smig.org.mx
Correo electrnico: administracion@smig.org.mx

Prohibida la reproduccin parcial o total de esta publicacin, por cualquier medio, sin la previa
Autorizacin escrita de la Sociedad Mexicana de Ingeniera Geotcnica, A.C.
Total or partial reproduction of this book by any medium requires prior written consent of the
Sociedad Mexicana de Ingeniera Geotcnica, A.C.
Las opiniones expresadas en este volumen son responsabilidad exclusiva de los autores.
Opinions expressed in this volume are the sole responsibility of their authors.

Collaborators (Editing and Formatting)/ Colaboradores (Edicin y Formato): A.R. Pineda Contreras,
E. Martnez Hernndez y A.L. Espinosa Santiago.
ii

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.


COUNCIL OF HONOR / CONSEJO DE HONOR
Leonardo Zeevaert Wiechers
Ral J. Marsal Crdoba
Alfonso Rico Rodrguez
Enrique Tamez Gonzlez
Guillermo Springall Caram
Edmundo Moreno Gmez
Carlos Jess Orozco y Orozco
Luis Vieitez Utesa
Gabriel Moreno Pecero
Ral Lpez Roldn
Ral Flores Berrones
Luis Miguel Aguirre Menchaca
Gabriel Auvinet Guichard
Luis Bernardo Rodrguez Gonzlez
Ral Vicente Orozco Santoyo
Alberto Jaime Paredes
Mario Jorge Orozco Cruz
Juan Jacobo Schmitter Martn del Campo
Hctor M. Valverde Landeros

CONSULTIVE COUNCIL / CONSEJO CONSULTIVO


Jos Francisco Fernndez Romero
Rigoberto Rivera Constantino
Walter Ivn Paniagua Zavala
Juan de Dios Alemn Velsquez
David Yez Santilln

BOARD / MESA DIRECTIVA 2015-2016


Ral Aguilar Becerril
President
Norma Patricia Lpez Acosta
Vice-President
Carlos Roberto Torres lvarez
Secretary
Celestino Valle Molina
Treasurer
Mara del Carmen Suarez Galn
Nilson Contreras Pallares
Miguel Figueras Corte
Aristteles Jaramillo Rivera
Technical Assistants
iii

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.


ORGANIZING COMMITTEE / COMIT ORGANIZADOR
3rd International Conference on Deep Foundations / 3er Simposio Internacional de Cimentaciones Profundas

iv

Walter I. Paniagua Zavala

ISSMGE

Juan Pauln Aguirre

ISSMGE

Norma Patricia Lpez Acosta

SMIG

Mary Ellen Large

DFI

Theresa Engler

DFI

Vernon Schaefer

G-I

2015

November 11-12th,
Mexico City

3rd International Conference on Deep Foundations


Deep Foundations and Soil Improvement in Soft Soils

vi

Foreword
On behalf of the Technical Committee TC-214 of the International Society for Soil Mechanics and
Geotechnical Engineering (ISSMGE), it is a privilege to present this volume for the 3rd International
Conference on Deep Foundations (Deep Foundations and Soil Improvement in Soft Soils), held in
Mexico City, November 11-12th, 2015.
This time, four organizations have joined efforts to produce it: the above mentioned TC-214
(Foundations Engineering for Difficult Soft Soil Conditions), the Mexican Society for Geotechnical
Engineering (SMIG, which hosts the TC-214), the Deep Foundations Institute (DFI), and the GeoInstitute of ASCE. In two previous events, SMIG and DFI had collaborated in 2011 and 2013 to
organize the First and Second International Conference on Deep Foundations, with very good
acceptance in the geotechnical community.
The purpose of merging different entities is multiple: to foster collaboration between countries, to
continue the technological and scientific knowledge transference, and to promote different points of
view from geotechnical professionals, including academicians, consultants, contractors and equipment
manufacturers.
Therefore, the material presented hereby, includes a wide spectre of the deep foundations and soil
improvement current knowledge, with special emphasis in soft soils. Three main topics are
recognized: Deep foundations, Excavations, and Soil Improvement. From state of the art of
geotechnical research to case histories, the papers presented herein give a general -and presentperspective on this matter.
My gratitude to all attendees, speakers, exhibitors and members of the Organizing Committee, for
their interest, collaboration and hard work in this event.

Walter I. Paniagua
TC-214, Chair
Pilotec, SA de CV

vii

viii

Introduction
The interest for constructing high-rise buildings in urban zones and the necessity to build structures in
difficult subsoil conditions requires engineers to look for efficient solutions. Scenarios are challenging
the projected structures that are affected by significant natural forces, such as those imposed by wind,
earthquakes or sea waves. The construction of deep foundations and the soil improvement works
have proved to be efficient alternatives to handle these situations.
In many cases, especially when loads and mechanical elements are of important magnitude, the design
of deep foundations is mandatory. For example deep foundations are employed when weak or
otherwise unsuitable soil exists near the subsurface and the vertical loads must be carried to depth
deposits. Deep foundations have other uses, for example they are used to resist scour, sustain axial
loading by side resistance in strata of granular soils or competent clay, allow above-water
construction, support lateral forces, improve the stability of slopes, reduce settlements and other
special purposes. The most used deep foundations are driven piles and drilled shafts.
In other cases, when the resistance or deformability conditions of soils are not allowable for the
project, the use of techniques for soil improvement are required. They help to reduce total or
differential settlements, increase axial and lateral bearing capacity and, in some cases, help to avoid an
undesirable soil behavior, such as liquefaction, swelling, among others.
The 3rd International Conference on Deep Foundations (3rd ICDF), held for the third time in Mexico
City, is a space to present recent experiences related to deep foundations and in this occasion it
includes the topic of massive improvement in soft soils. The aim of this conference is to promote the
most recent technical and scientific developments and to share experiences in the design and
construction of deep foundations and improvement techniques of soils.
The papers received for the 3rd ICDF include subjects such as cases history, foundations for high-rise
towers, geo-construction techniques, special deep foundations, cases of improvements on different
subsoils and deep excavations, among others. There is no doubt that the lectures on these topics will
also increase our knowledge of soil behavior.
My sincere acknowledgment to all authors for their invaluable contributions as well as to the
Organizing Committee for their efforts to achieve a successful Conference.
The Mexican Society for Geotechnical Engineering is proud to hold the 3rd ICDF on Mexico City
and will collaborate continuously with the Deep Foundations Institute, the Geo-Institute and the
International Society for Soil Mechanics and Geotechnical Engineering through its ISSMGE
Technical Committee TC-214, in order to promote the dissemination of the geotechnical knowledge.

Ral Aguilar Becerril


Presidente SMIG Mesa Directiva 2015-2016

ix

3rd

International Conference on Deep Foundations


Deep Foundations and Soil Improvement in Soft Soils

Technical Committee

TC-214

Contents
Page
Foreword............................................................................................................................

vii

Introduction........................................................................................................................

ix

SESSION 1. DEEP FOUNDATIONS

Effects of varved deposit on driven piles at a LNG terminal site


Efectos de depsitos estratificados en pilotes hincados en el sitio de la terminal LNG
LIN Guoming, HUANG Yanbo & LIN Cheng..............................................................
Low noise and low vibration press-in piling method in soft soil in congested urban
areas
Mtodo de piloteo de baja presin de vibracin y bajo ruido en suelos blandos en reas
urbanas congestionadas
TAKUMA Takefumi.....................................................................................................
The use of displacement piling technology in soft soil conditions
El uso de tecnologa de pilotes de desplazamiento en condiciones de suelo blando
MARINUCCI Antonio & CHIARABELLI Marco.........................................................
Rescate de una cimentacin de pilas con inclusiones rgidas
Pile foundation retrofit with rigid inclusions
SEGOVIA Jos, PANIAGUA Walter y LPEZ Germn...............................................
Foundation design and construction for high-rise Towers in Mexico City
Diseo de la cimentacin y construccin de Torres de gran altura en la Ciudad de Mxico
DEMING Peter W., NIKOLAOU Sissy, POLETTO Raymond J. &
TAMARO George J.....................................................................................................
Deep foundations in Mexico City soft soils
Cimentaciones profundas en suelos blandos de la Ciudad de Mxico
AUVINET-GUICHARD Gabriel & RODRGUEZ-REBOLLEDO Juan-Flix..............
The use of micropiles technology in soft soil conditions
El uso de tecnologa de micropilotes en condiciones de suelo blando
PAGLIACCI Federico.................................................................................................

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

11

19

31

39

51

67

xi

3rd International Conference on Deep Foundations

Contents

Deep Foundations and Soil Improvement in Soft Soils

Page
Pore pressure build-up due to pile driving in clayey deposits
Desarrollo de presin de poro debido al hincado de pilotes en depsitos arcillosos
MENDOZA
Manuel
J.,
RUFIAR
Miguel,
IBARRA
Enrique
&
OROZCO Marcos...........................................................................................................
Geotechnical design of the foundation for an office building located at the transition
zone
Diseo geotcnico de la cimentacin para un edificio de oficinas localizado en la zona de
transicin
ARENAS Fernando & CUEVAS Alberto.........................................................................

SESSION 2. EXCAVATIONS

77

85

91

A historic capitol and a deep excavation


Un capitolio histrico y una excavacin profunda
MASSOUDI Nasser & SLIWOSKI Richard....................................................................
The support of a 25 m deep excavation in difficult ground conditions using Single
Bore Multiple Anchor technology
Soporte de una excavacin de 25 m de profundidad en condiciones de terreno difcil
usando tecnologa de anclaje mltiple con barreno nico
MOTHERSILLE Devon & OKUMUSOGLU Bora.........................................................
The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground
seismic areas
El uso de muros MSE rellenados con concreto celular ligero en reas ssmicas de terrenos
blandos
PRADEL Daniel & TIWARI Binod.................................................................................

107

SESSION 3. SOIL IMPROVEMENT

115

Soil improvement around the world Applications and solution examples


Mejoramiento de suelos alrededor del mundo Aplicaciones y ejemplos de solucin
GERRESSEN F...............................................................................................................
Principles and application of soil mixing for ground improvement
Principios y aplicacin de la tcnica soil mixing para mejoramiento de suelo
WILK Charles M.............................................................................................................
Sustitucin dinmica aplicada en turbas de la pennsula de Yucatn
Dynamic replacement soil improvement technique applied in peaty soils in the peninsula
of Yucatan
CIRION ARANA Alfredo, CHATTE Rmi & PAULN AGUIRRE Juan.........................
Transforming marginal land to support a world class development in Panama
Modificacin de suelos marginales para apoyar proyectos de clase mundial en Panam
LANGONI Gustavo & ARCHABAL Roger.....................................................................

xii

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

93

97

117

123

127

131

3rd.

International Conference on Deep Foundations


Deep Foundations and Soil Improvement in Soft Soils

Session 1:
Deep foundations

Technical Committee

TC-214

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Effects of varved deposit on driven piles at a LNG terminal site


Efectos de depsitos estratificados en pilotes hincados en el sitio de la terminal LNG
Guoming LIN1, Yanbo HUANG2, and Cheng LIN 3
1

Senior Consultant, Terracon, 2201 Rowland Avenue, Savannah, GA 31404; Guoming.Lin@terracon.com


Geotechnical Engineer, Terracon, 2201 Rowland Avenue, Savannah, GA 31404; Yanbo.Huang@terracon.com
3Project Geotechnical Engineer, Terracon, 2201 Rowland Avenue, Savannah, GA 31404; Cheng.Lin@terracon.com

2Staff

ABSTRACT: More than 6000 steel pipe piles are required for a proposed large LNG (Liquefied Natural Gas) Terminal
project. The regasication facility includes three 3.5 billion cubic feet (100 million cubic meters) double-wall storage
tanks for a daily send-out capacity of 1.2 billion cubic feet (33 million cubic meters) and a pier designed to berth ships
with a capacity of 200,000 cubic meters. The subsurface conditions at the tank locations were explored with a
combination of 12 soil test borings (STB), 34 cone penetration test (CPT) soundings and eight dilatometer test (DMT)
soundings. The geotechnical study also included eld vane shear testing, pore pressure dissipation testing and
laboratory testing. The site subsurface conditions feature a layer of 90-foot thick very soft to stiff clayey silts with
interbedded thin sand seams (varved deposit). Characterization of this varved deposit layer, especially its shear strength
and preconsolidation history, is critical to the foundation design and construction for this project. However, the unique
structure of the clayey silts presents difficulties in dening some of its properties such as time rate of consolidation and
undrained shear strength. This paper presents the subsurface exploration program and the methods used to characterize
the clayey silts from the eld and laboratory testing results. The preconsolidation history of this layer was evaluated using
several different approaches. This paper discusses the potential downdrag force and its implication in the pile design and
construction. A statistical procedure is developed to analyze axial pile capacities using SPT and CPT based methods and
pile capacities obtained from different methods are compared and discussed.

1 INTRODUCTION
1.1 Project information
The proposed LNG Terminal includes three 150,000
cubic meter (944,000 barrel) double wall insulated
LNG storage tanks, process equipment consisting of
compressors and vaporizer, buildings, pipelines,
impoundment dikes, roads, and a parking lot. A jetty
will be built for unloading LNG tankers and a
breakwater may be built to provide a sheltered area
for the tanker. The tanks are designed to store
liqueed natural gas (LNG) at a pressure of 2.0 psig
and a temperature of -270F. The tanks will have an
outer concrete wall (122 feet in inner radius) and an
inner steel tank. The project will require a permit
from the Federal Energy Regulatory Commission
(FERC).
1.2 Site description
The site is located on the southeast bank of the
Delaware River in Logan Township, Gloucester
County, New Jersey. The property, approximately
175 acres, is primarily an agricultural soybean field

with several gas and liquid petroleum pipe lines that


traverse the Delaware River and make landfall on the
northwestern end of the property. The site is
generally at and had been used for disposing of
dredge spoil from the Delaware River before the
1960s.
1.3 Geotechnical testing
The subsurface conditions of the site were explored
with a combination of 12 soil test borings (STB), 34
cone penetration test (CPT) soundings, and eight
dilatometer test (DMT) soundings. Field vane shear
tests were performed at three STB locations. In
conjunction with the CPT soundings, pore pressure
dissipation tests were performed at various depths
within four CPT locations were measured at the
center of the three tanks. The laboratory testing
program consisted of soil index testing, consolidation
and triaxial shear strength testing, and chemical
analyses.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Effects of varved deposit on driven piles at an LNG terminal site

1.4 Subsurface Stratigraphy


The subsurface stratigraphy is generalized in the
following table.

sizes and thickness. The CPT soundings take


readings at 2-centimeter intervals, which is relatively
accurate in determining the interfaces of the soil
strata. Figure 2 is a contour map showing the bottom
elevation of the clayey silt layer varying approximately
from -86 to -93 feet (NAVD) under Tank 1.

Table 1. Generalized Subsurface Stratigraphy.


Layer
No.

Soil Type

Dredged Fill

2
3

Sand/Gravelly Sand
Clayey Silt with
Interbedded Sand
Seams
Sand and Gravel
Residual Clayey
sand

4
5

Average
Elevation
(ft, NAVD)
7 to 3

H*
(ft)

Geologic
Period

Recent
Dredge Spoil
Quaternary
Quaternary

3 to -3
-3 to -93

6
90

-93 to -113
-113 to -148

20
35

6
Metamorphic Rock
Below -148
*H is the average thickness of soil layer

Quaternary
Tertiary and
Cretaceous

--

Layer 3, termed varved deposit, has a thickness of


85 to 95 feet and contains clayey silts with numerous
interbedded thin ne sand seams as shown in Figure
1. Geologically, the soils were deposits of recent age
as a result of warmer temperatures and rise of ocean
levels after the Glacial Period. The ne grained soils
became fertile ground for vegetation which resulted
in variable amounts of organics within this layer. Due
to its great thickness, the shear strength of this layer
can greatly affect the pile capacities. Furthermore,
potential downdrag force is a concern if the layer is
underconsolidated or will undergo additional
settlements from the surface loads.

Figure 2. Bottom of the Varved Clay Layer.

2 CHARACTERIZATION OF THE VARVED


DEPOSIT (LAYER 3)
2.1 Soil index properties and classication
The Layer 3 soils are mostly classied as low
plasticity silts (ML) and high plasticity silts (MH) with
occasional classication of high plasticity clays (CH)
or clayey sands (SC). Table 2 summarizes the soil
index and classication properties.
Table 2. Summary of Soil Index and Classication
Properties.

Range
Average

# 200
Passing
(%)

Natural
Moisture
Content
(%)

Liquid
Limit

Plastic
Limit

Organic
Content
(%)

50~98
83

20~90
55

27~94
54

12~50
28

1.7~7.1
3.7

2.2 Consistency

Figure 1. Photo of Clayey Silt with Interbedded Sand


Seams.

SPT blow count, CPT tip resistance, and DMT


modulus were used to characterize the consistency
of the clayey silt. In general, the clayey silt exhibited
slightly increasing consistency with depth from soft at
the top to firm near the bottom.

Samples taken from the SPT samplers and Shelby


tubes allow visual observations of characteristics of
the sand seams, such as the depth intervals, particle
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

LIN G. et al.

can be considered to increase from 400 psf at the top


of clayey silt to 1100 psf at the bottom of the clayey
silt with an average value of 750 psf.

Figure 3. Unit Weight with Depth.

2.3 Unit weight

Figure 4. Undrained Shear Strength versus Depth


(Exponential Regression).

As shown in Figure 3, the unit weight derived from


DMT soundings agrees reasonably well with the
results of laboratory tests performed on Shelby
tubes. The total unit weights were found to increase
slightly with depth in the clayey silt layer from 100 pcf
at the top to 110 pcf at the bottom.

2.5 Compressibility

2.4 Undrained shear strength


The undrained shear strengths of the clayey silts
were obtained using three different methods:
laboratory triaxial tests, eld vane shear tests, and
correlations from CPT data. Due to the effect of thin
sand seams within the test specimens, almost all
unconsolidated undrained (UU) triaxial tests resulted
in a sloped failure envelope rather than a horizontal
failure envelope typical for normally consolidated
clays. The undrained shear strength was then
interpreted as the shear stress corresponding to the
effective in-situ overburden stress on the failure
envelope. The uncorrected undrained shear strength
obtained from eld vane shear tests was
approximately two times as large as the undrained
shear strength from the UU triaxial tests. Using the
Correction procedures of Bjerrum (1972) as revised
by Aas et al. (1986), the corrected vane shear
strength values agree reasonably well with the
laboratory test results, as shown in Figure 4. The
trend of undrained shear strength increasing with
depth can be approximated exponentially or linearly.
In CPT soundings, the shear strength is related to
the cone tip resistance by a cone factor Nkt. An Nkt
value of 15 to 18 was obtained by matching the
undrained shear strength derived from the CPT data
to the best tted curve from Figure 4. Previous
studies by others indicated the cone factor Nkt
generally ranges between 15 and 20 (ESOPT 1974
and 1982, ISOPT 1988). Undrained shear strength

Laboratory consolidation tests were performed using


both conventional incremental loading procedures
(ASTM D-2435) and constant strain rate (CSR)
method (ASTM D-4186). The conventional
consolidation tests yielded an average compression
index (Cc) of 0.566 with an average initial void ratio
(eo) of 1.535, which corresponded to an average
compression ratio [Cc/(1+eo)]of 0.223. The CSR
consolidation tests measured an averaged
compression ratio of 0.218. Constrained modulus of
compression (M), derived based on the empirical
correlations with DMT data (Schmertmann, 1988)
and CPT data (Senneset et al., 1989 and Kulhawy
and Mayne, 1990), varied approximately between 25
and 60 tsf for the clayey silts. The constrained
modulus derived from the three CSR consolidation
tests averaged about 38 tsf.
2.6 Time rate of consolidation
The time rate of consolidation was characterized
using laboratory consolidation tests and in-situ CPT
pore pressure dissipation tests. Theoretically, the
drainage path for the laboratory consolidation tests is
in the vertical direction while the pore pressure
dissipation tests measure pore pressure dissipation
in the horizontal direction. The vertical coefficient of
consolidation (CV) measured in the consolidation
test, varied between 0.1 and 0.5 ft2/day around the
in-situ overburden stress. The coefcients of
consolidation in the horizontal direction, Ch, were
calculated based on a method proposed by Mayne
(2002). The Ch, values varied from 0.41 ft2/day to
245 ft2/day. The large variation of the Ch values

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Effects of varved deposit on driven piles at an LNG terminal site

suggests that the rate of pore pressure dissipation


(i.e. consolidation) is greatly affected by the
interbedded sand seams within the clayey silts
3 PRECONSOLIDATION HISTORY OF THE
VARVED DEPOSIT (LAYER 3)
The consolidation history of the varved deposit, i.e.,
whether the clayey silts are underconsolidated,
normally consolidated or overconsolidated, is
essential for the foundation design of this project.
This condition will affect the magnitude of settlement
of lls and shallow foundations and whether to
consider downdrag forces on deep foundations. Due
to the great thickness of the clayey silt layer and
large number of piles to be used on this project, this
consideration has substantial impact on the design of
this project.
Geologically, the clayey silts (varved deposit) were
deposited along the shores of the Delaware River
during the late Pleistocene Period. The deposition of
the layer was a long slow process taking place more
than ten thousand years ago. It is logical to assume
that the silts and clays had consolidated under the
self-weight of the material. The dredged fills were last
deposited at the site in the 1960s. From the
geotechnical standpoint, the clayey silts are more
than 80 feet in thickness, which would require a long
time to consolidate. However, the interbedded sand
seams would function as horizontal drainage paths to
facilitate consolidation. As such, the key question is if
the consolidation of the clayey silts has completed
under the weight of the dredged ll placed more than
30 years ago. Several different approaches were
taken in evaluating the preconsolidation history of the
clayey silt layer at the site.

curve using linear regression intercepts the strength


axis at 75 psf (very small). The average strength gain
is 15.9 psf per foot. Dividing the average strength by
the average effective vertical stress gives a c/p ratio
of around 0.37, a reasonable value for a normally
consolidated or slightly overconsolidated silt.
3.3 Overconsolidation ratio
In laboratory consolidation tests, smaller load
increments were added in the vicinity of the existing
overburden pressure to ne-tune the compression
curves for the determination of preconsolidation
pressure. The preconsolidation stresses, determined
using the Casagrande procedures, indicated OCR
values ranging between 0.9 and 2.1 with an average
of 1.26. The OCR values were also derived from both
CPT and DMT using empirical correlations (Powell
and Cuarterman, 1988; Kamei & Iwasaki, 1995). The
OCR of the clayey silts derived from DMT and
obtained through consolidation tests fall mostly
between 1 and 3, as shown in Figure 5. Therefore,
the clayey silts are considered normally to slightly
over-consolidated based on the overconsolidation
ratio

3.1 Atterberg limits


The relationship between the natural moisture
contents and Atterberg Limits can be used as an
approximate indication of soils preconsolidation
history. Moisture contents that are well above the
liquid limit at depths of tens of feet usually indicate
underconsolidation. Moisture contents near the
plastic limit at shallow depths usually indicate
overconsolidation. A statistical analysis of the
laboratory test results performed on 30 Shelby tube
samples indicated an average natural moisture
content of 55 percent and an average liquid limit of
54 percent. The natural moisture contents are very
close to the liquid limits. These properties lead to the
conclusion that the soil is not significantly
overconsolidated or underconsolidated.
3.2 Undrained shear strength
The undrained shear strengths generally increase
with depth (effective vertical stress). The best fit

Figure 5. OCR from DMT and Consolidation Test.

3.4 Consolidation theory


The Cv and Ch values obtained from the laboratory
consolidation tests and derived from the eld pore
pressure dissipation tests vary greatly with depths
and locations. The thickness of clay layer between
two drainage paths (sand seams) also vary greatly.
Therefore, a conservative model was used to predict
the time required for the consolidation of the clayey
silts under the weight of the dredged ll. Using the
Terzaghis one dimensional consolidation theory and
a Cv of 0.01 ft2/day and a layer thickness of 10 feet
between top and bottom drainage paths (both values
are considered conservative based on the eld and

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

LIN G. et al.

laboratory test results), a time of 5.8 years is required


to reach 90 percent consolidation. Considering the
dredged ll was last placed on site more than 30
years ago, it is reasonable to believe that the clayey
silts had consolidated under the weight of the
dredged ll.
4 PILE DESIGN CONSIDERATIONS
Steel pipe piles without concrete ll are considered
the most suitable foundation system for this project
(Yang et al., 2003). All piles are required to be
embedded into the sand and gravel (Layer 4).
Preliminary design performed by the design engineer
(CB&l) requires an ultimate axial compression
capacity of 200 tons and 280 tons for 18 and 22 inch
diameter piles, respectively.

based on the factor recommended by American


Petroleum Institute (API, 1993). For the clayey silt
layer dened in the Driven program, undrained shear
strength of 500 and 1000 psf, corresponding to
adhesion values of 400 and 800 psf based on
Tomlinsons (1980) method, were used as the lower
and upper bound values, respectively.
Figure 8 presents comparison of the ultimate axial
capacities using the two CPT based methods versus
the Driven program. For the CPT based methods, the
average pile capacities obtained through the
statistical procedure are presented. It appeared the
ultimate axial pile capacities calculated from the
three different methods agree relatively well with
each other. Back-calculation based on the pile
capacity vs. depth curves indicated the average
adhesion values of the clayey silt were 500 and 750
psf for the French Method and the Eslami & Fellenius
Method, respectively.

4.1 Axial Pile capacities


0

Eslamic & Fellenius (1997) Direct CPT Method,


10 to 90% of Normal Distribution

20

40

Depth (ft)

The axial capacities will be largely dependent on


the adhesion between the pile and the clayey silt
layer (Layer 3). Several methods were used to
estimate the axial capacities of driven piles under
static loads. These methods included calculating
adhesion and friction values between pile and soil
based on the soil strength from laboratory and eld
testing, directly from the CPT results, and based on
SPT blow counts using a computer program FHWA
Driven 1.2 (2001).
Cone penetration tests with pore pressure
measurements (CPTu) are considered probably the
best in-situ test method for the design of axially
loaded piles (Hannigan, et al. 1997). Various
calculation methods based on CPT data were
reviewed and compared, and two methods were
selected to estimate axial pile capacities for this
project: the method developed by Bustamante and
Giasenelli (1982), also called the LCPC method or
the French method and the method proposed by
Eslami and Fellenius (1997). Before the calculation,
the depths associated with CPT data were adjusted
to a ground surface at 6 feet NAVD.
To account for the variations of the soil conditions
at this site, a statistical analysis was performed by
assuming pile capacities based on the CPT at
different soundings would have a normal distribution
at the same depths. The procedure generated an
upper bound, lower bound and an average pile
capacity at a given depth by eliminating the samples
which signicantly deviate from the main group (more
specically, values of probability density function less
than 10 percent). Figures 6 and 7 present the
ultimate pile axial capacities for 18-inch steel pipe
piles calculated using the two CPT methods and the
statistical procedure.
An average undrained shear strength of 750 psf
was derived from the laboratory and field tests, which
corresponds to an adhesion of approximately 656 psf

60

80

100

120
0

50

100

150

200

250

300

350

400

450

500

Ultimate Compression Capacity (tons)

Figure 6. Ultimate Compression Capacities for 18-in Steel


Pipe Pile Using CPT methods by Eslamic & Fellenius
method.

An ultimate compression capacity of 200 to 280


tons was recommended for 18-inch and 22-inch OD
pipe piles, respectively, after the piles are driven into
the sand and gravel layer. The pile length will vary
with elevations of the top of the sand and gravel layer
as well as the depth of penetration into this layer. A
pile tip elevation between -95 and -108 feet NAVD
can be expected for the purpose of preliminary
design and estimates.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Effects of varved deposit on driven piles at an LNG terminal site

the piles to be embedded into the sand and gravel


layer (Layer 4) even though the piles may achieve
the required pile capacities with the pile tips above
this layer.
5 SUMMARY AND CONCLUSIONS

Figure 7. Ultimate Compression Capacities for 18-inch


Steel Pipe Pile by French Method.

1. The subsurface conditions for the proposed BP


Crown Landing LNG terminal site were explored by
using a combination of 12 soil test borings, 34 cone
penetration test soundings and eight dilatometer
soundings. The site features a thick clayey silt layer
with interbedded fine sand seams. The strength and
compressibility characteristics of this clayey silt layer
have a major impact on the costs and safety of the
pile foundation. The soil properties were tested using
a series of eld vane shear tests, pore pressure
dissipation tests and laboratory tests.
2. The combined use of SPT, CPT and DMT
soundings was a well thought-out choice for
subsurface exploration. The samples from the SPT
samplers and Shelby tubes allowed the engineers to
closely examine the characteristics of the
interbedded layers, such as the depth intervals,
thickness and particle sizes of the sand seams. CPT
soundings provided more accurate determination of
the depth of soil interface and continuous data for
deriving other engineering properties and subsequent
pile capacity calculations.

Figure 8. Comparison of Compression Capacities by


Different Methods.

4.2 Downdrag Force


The preliminary grading plan shows the nal grade in
the tank areas will be at or near the existing site
grade. There will be minimal increase of stresses in
the soils from the fill and grading in the tank area.
Based on analyses presented in the previous section,
it was concluded that the clayey silts are normally
consolidated. No downdrag force was considered for
the piles underneath the tanks. However, there are
uncertainties associated with potential secondary
consolidation and long-term decomposition of
organic materials in the soils. The risk associated
with the above uncertainties is considered relatively
small. For the piles embedded in the sand and gravel
layer (Layer 4), there is an extra 100 to 150 ton
compression capacity available to offset this highly
unlikely but potential downdrag force. The
Construction quality control program will require all

3. The undrained shear strength of the clayey silts


was determined using three different methods: eld
vane shear tests, laboratory triaxial tests and
empirical correlations from the CPT soundings. The
results from different methods agree reasonably well
after proper corrections.
4. The consolidation history of the clayey silts was
analyzed qualitatively based on the geological and
geotechnical considerations and quantitatively using
laboratory and eld test results. The analyses
consistently indicate that the clayey silts are normally
consolidated or slightly overconsolidated. No
downdrag force was considered necessary for piles
supporting the tanks, where no stress increase from
ll or grading is anticipated.
5. Axial pile capacities were calculated using three
different methods: based on adhesion and friction
values of the soils, directly calculated from the CPT
data and SPT blow count values. Results from these
methods were evaluated and compared. Cone
penetration test data was considered the best
method in calculating pile capacities. A statistical
analysis procedure was developed to account for the
variations of the soil conditions to present the pile
capacities in upper and lower bounds.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

LIN G. et al.

6 ACKNOWLEDGEMENTS
The authors would like to acknowledge the following
organizations and individuals for their help and
support for this project: BP and Mr. Junius Allen,
CB&I and Messrs Greg Bertha and Donald Barrs, Dr.
Felix Yokel, Site Blauvelt Engineers, Geotesting
Express,
and
many
current
and
former
Terracon/WPC colleagues Messrs. William Wright,
Wu Yang, Edward Hajduk, Jian Fang, Thomas
Casey, William Anderson and Donovan Ledford.

Powell, J. J. M., & Quarterman, R. S. T. (1988). The


Interpretation of Cone Penetration Tests in Clays,
with Particular Reference to Rate Effects,
Penetration Testing, Balkema, Rotterdam, The
Netherlands, Vol. 2: 903-909.
Tomlinson, M. J. (1980). Foundation Design and
Construction, Pitman, London, UK
Yang, W., Fang, J., and Lin, G.M. (2003). WPCs
Geotechnical Report Liberty LNG Project.
Savannah, Georgia:

REFERENCES
Aas, G., Lacasse, S., Lunne, T., & Hoeg, K. (1986).
Use of In-situ Tests for Foundation Design on
Clay, Publikasjon-Norges Geotekniske Institutt,
(166): 1-15.
American Society for Testing and Materials (2004).
Standard Test Methods for One-Dimensional
Consolidation
Properties
of
Soils
Using
Incremental Loading, ASTM D-2435.
American Society for Testing and Materials (2006).
Standard Test Method for One-Dimensional
Consolidation Properties of Saturated Cohesive
Soils Using Controlled-Strain Loading, ASTM D4186.
API (1993). Recommended Practice for Planning,
Designing and Constructing Fixed Offshore
Platforms Working Stress Design, 20th Edition,
American Petroleum Institute, Washington, DC.
Bjerrum, L. (1972). Embankments on Soft Ground,
Proceedings of Performance of earth and earthsupported structures, ASCE. 1-54.
Bustamante, M., & Gianeselli, L. (1982). Pile
Bearing Capacity Prediction by Means of Static
Penetrometer CPT, Proceedings of the Second
European Symposium on Penetration Testing.
493-500.
Eslami, A., & Fellenius, B. H. (1997). Pile Capacity
by Direct CPT and CPTu Methods Applied to 102
Case Histories, Canadian Geotechnical Journal,
Vol.34(6): 886-904.
Federal Highway Administrations Driven 1.2
computer program (2001). Blue-Six Software, Inc.
Hannigan, P. J., Goble, G. G., Thendean, G., Likins,
G. E., & Rausche, F. (1997). Design and
Construction of Driven Pile Foundations-Volume I
& II, FHWA-HI-97-014, Washington, D.C.
Kamey, T. & Iwasaki, K. (1995). Evaluation of
Undrained Shear Strength of Cohesive Soils Using
a Flat Dilatometer, Soils and Foundations, Vol
35(2): 111-116.
Mayne, P. W. (2002). Equivalent CPT Method for
Calculating Shallow Foundation Settlements in the
Piedmont Residual Soils Based on the DMT
Constrained Modulus Approach from http://www.
geosystems.ce.gatech.edu/~geosys/.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Low noise and low vibration press-in piling method in soft soil in congested
urban areas
Mtodo de piloteo de baja presin de vibracin y bajo ruido en suelos blandos en reas urbanas
congestionadas

Takefumi TAKUMA1
1Giken

America Corp.

ABSTRACT: Pile driving causes ground vibrations and noise by pile driving machines used during
construction. Conventional pile driving has all but disappeared from urban construction because of the
emissions of deafening noise and earth shattering vibrations. The Press-in Piling Method was developed to
solve most of the problems associated with pile driving in urban construction. This innovative piling technique
allows sheet piles and other types of prefabricated piles or panels to be hydraulically pressed into the ground
using the reaction force generated by the previously installed piles surface friction with soil and the systems
own weight. The Press-in pile driving machine walks on top of the pile wall, gripping on previously installed
piles while installing the next pile immediately adjacent to the one just installed. While the method is highly
suited for soft ground, the system can also efficiently deal with hard soil, such as dense sand, stiff clay,
gravels, cobbles, boulders and soft rock with attachments without another set of large equipment for predrilling. Some of the urban projects require pile driving in low head room or with very small clearance from
existing structures. In other cases, the pile driving may have to be conducted without an access road to the
piling location. This paper presents as to how the Press-in Piling Technology can effectively mitigate the
negative environmental impacts associated with pile driving on urban infrastructure projects along with case
studies in Japan, U.S.A. and Mexico.

1 INTRODUCTION
The Press-in Piling Method was invented out of pure
necessity back in 1975 in Japan, where very many
infrastructure projects were simultaneously built
nationwide due to the countrys rapid economic
expansion as well as the governments policy at that
time. A sheet pile driving project in a regional city of
Kochi, some 500 kilometers west of Tokyo, was
forced to shut down due to a noise and vibration
complaint filed by a local resident who lived right
next to the project. He had to rest during the daytime
due to his night time work. This incident prompted
Akio Kitamura, who was the president of the local
foundation contractor involved, to start thinking
about an alternative pile driving method which would
not generate noise or vibration. By collaboration with
a local inventor who was dubbed Edison of Kochi,
he built the first Press-in piling machine. Although
the original purpose for creating the pile driver was
to utilize it for his companys sheet pile driving
projects, foundation contractors in other regions of
the country, who were also looking for a low noise
and low vibration pile driving method, started to ask

about the equipment. That was the beginning of the


success story of the Press-in Piling Method. By now,
it has been widely used not only in Japan but also in
many parts of Asia, Europe and North America,
providing environmentally-friendly solutions to
numerous foundation projects.
2 HOW DOES THE PRESS-IN PILING WORK?
The Press-in Piling Method typically utilizes reaction
force derived from a few previously installed piles to
hydraulically push the next pile into ground (see
Figure 1). The Press-in piling equipment of this type
grips the top of already-driven piles to drive the new
one and moves forward or backward on its own
(See Figure 2). Due to the fact that this method is
not using vibrating or purcussive force to drive the
piles, it is regarded as a very environmentally-friedly
piling method.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

12

Low noise and low vibration press-in piling method in soft soil in congested urban areas

Figure 1. Principle of the Press-in Piling Method.

Low noise and practically vibration free.


The equipment size is relatively small and its
clamping points on the pile are much lower
than those of other piling methods. This
enables the equipment to work in physically
tight working conditions, horizontally and
vertically.
With attachments it can effectively drive piles
into hard soil.
It can achieve much more accurate pile
installation thanks to a combination of the
better control of the pile and the lower
clamping points compared to other piling
methods.

Its advantages are;

Figure 2. Sequence of the Press-in Pile Driving.

3 PRESS-IN PILE DRIVING IN HARD SOIL

3.1 Water Jetting


Dense sand, stiff clay, gravel, cobbles and boulders
are difficult to drive piles into. In some cases,
limestone, mudstone and weathered rock layers
may exist in pile lines. The high pressure water jet
and crush auger attachments as part of the Press-in
piling technology are very useful tools to drive piles

into some of these hard soil. The high pressure


water jetting will be quite effective in dense sand and
silt layers. A small nozzle attached to the toe of each
pile blasts out a small diameter of high pressure
water to create a pilot hole in a hard soil, loosen it
and lubricate the pile surface, reducing piles skin
friction. See Figure 3.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

TAKUMA T.

13

Figure 3. Press-in Piling with Integrated High Pressure Water Jetting for Dense Sand and Silt.

3.2 Crush Auger Attachment


The integrated auger, which simultaneously drills
into hard soil as the pile is pressed in, allows the pile
to be advanced by loosening, crushing and partially

removing the hard soil to accomplish the smooth pile


driving. Figure 4 shows how the auger system
works.

Figure 4. Press-in Piling with Integrated Crush Auger Attachment for Hard Soil.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

14

Low noise and low vibration press-in piling method in soft soil in congested urban areas

4 NON-STAGING PRESS-IN PILING FOR


PROJECTS WITH LIMITED ACCESS
The Press-in pile driving can be done by having all
the necessary equipment moving on top of alreadydriven piles. In other words, a line of driven piles can
be used as construction road by the Non-staging
Press-in Piling Method. Pile Runner transports the
piles from a faraway access point to the pile driver.

Clamp Crane picks up the pile from Pile Runner and


pitches it to the pile driver. See Figure 5. The
method is highly useful for projects that have limited
access, such as those above shallow or deep water,
ones on a sloped embankment without the need for
staging or dense jungle without any construction
road.

Figure 5. Non-staging Press-in Piling Method with Clamp Crane and Pile Runner.

5 CASE STUDIES
5.1 Myoshoji River Restoration Project (Gekitoku-1
Section), Tokyo, Japan
Myoshoji River is one of Kanda Rivers tributaries
located approximtely 15 km northwest of downtown
Tokyo. Although only 9.7 km long and its watershed
being relatively small, the river runs through densely
populated residential and commercial areas of the
city. Very heavy rainfall (263 mm) on one September
evening in 2005 flooded more than 3,300 units of
buildings in the area. To reduce such flooding in the
future, this project was to widen the river to increase
the drainage capacity and to reduce the flood risk by
installing 634 of 1,000 mm diameter tubular piles
into the existing concrete retainig walls. The pile
depth varied from 11 to 22 m. Rotary Press-in Piling
equipment was utilized to effectively drive the piles
into the concrete retaining walls without removing
them. Each pile had cutter bits welded at the toe of
the pile to facilitate the cutting operations. Due to the
roads on both sides of the river being quite narrow

for site access and also to the fact that they had to
be kept open for the local traffic almost all the time,
the Non-staging Press-in Piling Method was
adopted. The tube piles were delivered to the
projects material handling point by a flat bed truck
and tranferred to Pile Runner that subsequently
traveled on the rail placed on the pile top to Clamp
Cranes pick-up point without blocking the road
traffic.
Figure 6 shows the projects sectional view and
Figure 7 shows sites soil conditions containing
dense sand, sandy gravel and consolidated silt
layers with SPT values at or higher than 50 at 8 m
and continuously beyond 12 m below ground. The
Rotary Press-in Piling equipment used a small
quantity of water as lubricant for efficient rotary
cutting operations. Figure 8 shows the Rotary Pile
Driver at work just in front of a local clinic.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

TAKUMA T.

15

Figure 6. Sectional View of the Project.


Figure 8. Pile Driving in Densely-populated Area.

5.2 Sandalwood Canal Improvement Project


(Hodges Bl. from Beach Bl. to Atlantic Bl., Project
No. P-80-01) in Jacksonville, Florida, U.S.A.

Figure 7. Soil Conditions.

This project was to repair the damaged earthen


levees by an earlier flooding as well as to increase
the drainage capacity of an existing canal by
widening/deepening with sheet piles driven into the
levees running through a densely populated area of
the City of Jacksonville, Florida. In order to minimize
noise/vibration and also to reduce in-stream
exposure of the equipment during construction, two
units of the Press-in pile drivers were used and the
work was done during the dry season of winter. The
widths of levee shoulders were relatively narrow
(approximately 3 m, see Figure 9) for a truck crane
to maneuver through, so a 10 ton capacity Clamp
Crane was used for hoisting sheet piles to the pile
drivers. The soil conditions were primarily sandy with
the SPT values of between 10 and 45 as shown in
Figure 10. The noise and vibration during the sheet
pile driving were limited by the specifications in the
following manner.
The hydraulic press-in equipment shall not
produce more than 70dB of noise, at a distance of
25 feet from the equipment, while in operation. It
shall not produce any measurable vibration at the
ground surface, at a distance of 25 feet from the
equipment, while in operation.
Approximately 950 pairs of 7.0 m to 9.0 m long Zshaped PZC18 type sheet piles were driven without
causing damages to the nearby homes. Figures 11
and 12 show the jobsite before and during the sheet
pile driving. The channel was sandwiched by rows of
houses.
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

16

Low noise and low vibration press-in piling method in soft soil in congested urban areas

Figure 9. Typical Cross Section.

Figure 10. Soil Conditions.

Figure 11. Sandalwood Canal After Vegetation Removal.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

TAKUMA T.

17

Figure 12. Two Units of Press-in Pile Drivers With Clamp Crane.

5.3 Foundation Reinforcement Work of San Juan


de Ulua Fortress, Veracruz, Mexico

deal with the harder layer in order to drive 18 meter


long U-shaped LX32 sheet piles.

The 16th century fortress foundation needed to be


rehabilitated and also protected to allow nearby
canals widening construction. To protect and
reinforce the foundation, sheet piles were driven to
form permanent retaining walls outside the
perimeter of the fortress standing in sea water.
Vibration from pile driving had to be minimized by
any means not to damage this invaluable historical
landmark. The agency in charge decided to adopt
the Press-in Piling Method to achieve this goal.
Sheet pile alignments were on the west, south and
southeast sides of the fortress as shown on Figure
13.

San Juan de Ulua


Fortress

Figure 14. Boring Data.


Sheet Pile Line

Figure 13. Plan View of the Sheet Pile Wall.

Although the water depth was relatively shallow (3


to 4 meters) and the soil is generally soft, there were
some harder layers due to dense sand mixed with
shells as seen on Figure 14. The high pressure
water jet attachment was employed to effectively

Sheet piles were driven right against the fortress


foundation on east and west sides. There was one
Y-shaped connecting point with four sharp angle
corners of sheet pile walls. A barge-mounted crane
was used to pitch the sheet piles to the Press-in pile
driver working on top of the sheets as shown on
Figures 15 and 16.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

18

Low noise and low vibration press-in piling method in soft soil in congested urban areas

6 CONCLUSION
Conventional pile driving methods not only cause
hated noise and vibration but also are not useable
on many projects in congested and densely
populated cities due to various local conditions. On
the other hand, the Press-in Piling Method can
achieve the projects goal in harmony with such
urban environment. With attachments and auxiliary
systems, it has wide range of applications in highly
congested conditions. With ever growing population,
the author believes that major cities in the world
including Mexico City and other populous cities in
the country will be greatly benefitted by adoption of
this method. It has shown the ability to preserve the
nations historical landmark in Veracruz.
Figure 15. Section View of Sheet Pile Driving Work.

REFERECES
White, D., Finlay, T., Bolton, M., and Bearss, G.
(2002),
Press-in Piling: Ground Vibration and Noise
During Piling Installation,
Proceedings of the International Deep Foundation
Congress, ASCE Special Publication 116
Motoyama M., Goh, T. and Yamamoto, M. (2005)
Silent Piling Technology and Its Application in
Hong Kong, Proceedings of the 2005 conference
of the Hong Kong Branch of the Chartered
Institution of Highways and Transportation
Figure 16. Profile of Sheet Pile Driving Work.

The sheet piles were driven accurately without


causing damage to the foundation or the fortress
structure. The operator was operating the pile driver
on its own staging affixed to the machine. See
Figure 17.

Figure 17. Sheet Pile Driving at Southwest Corner of the


Fortress.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

The use of displacement piling technology in soft soil conditions


El uso de tecnologa de pilotes de desplazamiento en condiciones de suelo blando
Antonio MARINUCCI, PhD MBA PE1, & Marco CHIARABELLI1
1Soilmec

North America, Houston, Texas, USA

ABSTRACT: Displacement piles are cast-in-pace, reinforced concrete piles that are formed with little or no soil removal,
where the soil is displaced radially into the soil by the drilling tool. This piling technique is applicable in soft-to-firm ground
conditions in loose to medium dense sands and in cohesive soils where the undrained shear strength is less than about
100 kPa (2,000 psf). There are many benefits to the use of displacement piles, including low vibrations during pile
construction, minimal amount of soil removal, no need for stabilizing fluids (slurry), and improvement of the load resistance
especially in side friction. The benefits of displacement piling make this technology ideal in contaminated and/or urban
environments. This paper provides an overview of the various types of displacement piles that have been used, applicability
of the technology, and general requirements for types of equipment and tooling needed. In addition, practical examples of
the technology and recent advancements to displacement piling tooling will be presented as mini case histories.
.

1 OVERVIEW
Drilled displacement piles (DDP) refers to a
specialized technology in which a bored pile is
constructed using a process in which (1) a specially
designed tool is advanced into the ground using both
rotation and downward thrust (crowd force) to
displace the in situ soil radially outward into the
surrounding formation, and (2) concrete is injected
and steel reinforcement (if required) is inserted to fill
the created hole and provide structural stiffness. A
key benefit of DDP is the minimal amount of drill spoils
generated, which provides a cost effective and
practical solution for sites with contaminated soils
(e.g., typically found at landfills, brownfield sites, and
industrial facilities).
In addition to the reduced
environmental impact, other advantages of DDP such
as proven reliability, relatively rapid construction, high
daily production, minimal noise associated with DDP,
and minimal ground vibrations have contributed to the
increased use of the technique especially for
construction in urban areas, in congested spaces, and
in close proximity to existing structures.
DDP has been used as structural foundation
elements (e.g., support column loading) and for
ground improvement (e.g., column-supported
embankments) on both commercial and public work
type projects. The maximum diameter and depth that
can be achieved are directly related to the capability
of the drill rig used to construct the DDP. As reported
in the literature, displacement piles with diameters
ranging from about 300 to 800 mm (12- to 32-inches)
and to a maximum depth of approximately 35 m (115
ft).

1.1 Description and Classification


The myriad types of bored piles are typically classified
according to the qualitative amount of disturbance
resulting from the piles construction, which can range
from non-displacement type to a complete or full
displacement type of pile. Drilled shafts fall under the
non-displacement type of piling, and continuous flight
auger (CFA) piles can be categorized under either
non-displacement or partial displacement type
depending on whether (a) the concrete/grout is
injected under pressure, and (b) the ratio of the outer
diameter of the hollow drill stem to the diameter of the
borehole is greater than about 50-60%, in which case
a greater amount of soil will be displaced radially and
compacted into the borehole wall. That is, a narrower
hollow stem will result in minimal-to-partial
displacement of the soil, while a wider hollow stem will
result in a greater amount of soil being displaced into
the surrounding soil during drilling.
DDP can be listed under either partial displacement
or full displacement type according to the installation
method and/or the type/shape of tooling used to
create the pile, which can be grouped as essentially
cylindrical shaped (Figure 1a) or screw shaped
(Figure 1b).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

20

The use of displacement piling technology in soft soil conditions

1.2 Benefits provided by Displacement Piles


Various practitioners (Basu et al, 2010; Paniagua,
2006; Baxter et al, 2006; Bottiau, 2006; Brown, 2012;
NeSmith, 2004; Pagliacci and Chiarabelli, 2015; etc.)
have extolled the benefits realized through the use of
displacement piles, which include the following:

(a)

(b)

Figure 1. Schematic of (a) cylindrical shaped and (b) screw


shaped displacement piles.

A schematic of a representative displacement tool


is shown in Figure 2, which highlights many of the
common components found on modern DDP tools. In
general, modern displacement tools will contain the
following common elements (Figure 2): (1)
displacement body, which is an enlarged section near
the bottom of the drill string that facilitates soil
movement radially outward, thereby displacing it into
the surrounding soil, (2) a drilling tip attached to the
bottom of the drill string that is used to loosen the soil
during the advancement of the tool (if a re-usable
drilling tip is used, a pivoting gate located near the
bottom of the tool or drill string is utilized for the
injection of the concrete or grout; otherwise, with a
sacrificial drilling tip, the concrete or grout is injected
through the bottom of the drill string), (3) a hollow stem
drill string with a diameter smaller than or equal to the
diameter of the displacement body, (4) a lower auger
segment with partial flights that moves the soil upward
toward the displacement body, and (5) an upper auger
segment with partial flights that moves the soil
downward toward the displacement body.

Figure 2.
Schematic of a representative DDP tool
delineating many of the common components found on
most modern DDP tools (DeWaal displacement pile tool
shown; modified from Basu et al, 2010).

Environmentally friendly because minimal amount


of drill spoils produced return to ground surface,
thereby lowering both the risks associated with
transport of spoils (especially contaminated material)
and the cost of disposal;
Minimal vibration induced during the construction of
the displacement pile because the rotary
drilling technique does not induce large vibrations into
the soil;
Even in loose soils, the borehole can be formed
without need of steel casing and/or slurry;
Cleanness of the working platform, lowering the
risk to injury of onsite personnel;
Compared with non-displacement bored pile
techniques, the concrete overbreak is significantly
lower; and
Compared to non-displacement bored pile
techniques, higher unit side friction and end bearing
resistance can be achieved through the compaction of
the surrounding soil, which results in a lower cost (per
ton of load).
2 INSTALLATION-INDUCED CHANGES
During the construction, the soil surrounding the DDP
will undergo changes to its stress state (e.g., change
in void ratio) as a function of the soil type, original
stress state and consistency, shape of the tool, and
installation method. The changes are directly caused
by the loading imposed on the soil by the tool from
both radial compactive/torsional stresses and vertical
shearing stresses during the advancement and
extraction of the tool.
Ultimately, the ground
improvement induced by the installation process
results in larger unit values of side shear.
Consequently, the load-displacement response of a
displacement pile is comparatively stiffer than that of
a similarly sized non-displacement pile; therefore,
compared with similarly sized non-displacement piles,
DDPs will be able to achieve a given load resistance
at a shorter length.
In loose to medium-dense cohesionless soils, the
compactive effort of the tool produces greater radial
displacement and results in a decrease of the void
ratio (higher relative density than initial) relatively
soon after construction is completed. Brown (2005)
reported that DDPs increase the horizontal stress in
the ground and densify sandy soils around the pile
during installation[achieving] a measure of ground
improvement around the pile. For soft to stiff
(displaceable) cohesive soils, the soil will be deformed

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MARINUCCI A. et al.

plastically, may require some time to realize the


increase in shear strength as the soil undergoes
consolidation within the affected zone. In sensitive
soils, the disturbance caused by the displacement tool
during advancement and extraction could result in
remolding and formation of residual shear planes.
In partially saturated and fully saturated soils, the
advancement and extraction of the displacement tool
may generate excess pore water pressures in the soil
surrounding the pile. For cohesionless soils with
minor fines content (<15%), the dissipation of these
induced excess pore water pressures will be relatively
rapid. For cohesionless soils with appreciable fines
(>15%) and for cohesive soils, however, the
dissipation of the induced excess pore water
pressures will require time to dissipate, which will
depend on the length of the drainage path. Brown
(2012) warned that the construction of DDPs may
induce excess pore water pressures in the
surrounding soil, which could result in water intrusion
into a newly constructed pile as the excess pore
pressure dissipates.
3 APPLICABILITY
The technique is ideally suited for a wide spectrum of
soil conditions ranging from sandy gravel to clay, with
the caveat that the soil is able to be both displaced
and compacted. In cohesionless soils, displacement
piles are appropriate in loose to medium-dense soil
conditions, where the relative density (Dr) is less than
about 65%, the cone tip resistance (qc) is less than
about 14 MPa (2,000 psi), and/or SPT N-values are
less than 30 blows/0.3 m (or 30 blows/ft). Due to the
compactive nature of the displacement tool, the void
space is decreased, the structure is reorganized, and
the relative density increased, which has a positive
effect on the behavior of the DDP. NeSmith (2002)
reported that the installation of displacement piles in
dense cohesionless soils (qc greater than 14 MPa or
2,000 psi) can be difficult and uneconomical.
In cohesive soils, displacement piles are
appropriate in soft to stiff soil consistency, where the
undrained shear strength does not exceed about 100
kPa (2,000 psf) or where SPT N-values are less than
about 10 blows/0.3 m (or 10 blows/ft). During
installation, the cohesive soils undergo plastic
deformation and are compacted. Stiff cohesive soils,
however, are difficult to compact. Brown (2012)
indicated that residual soil profiles, weak limerock
formations, cemented sands, and stiff clays are soil
types that favor easy construction of displacement
piles. Reporting on conditions in the United Kingdom,
Baxter et al (2006) described that applicable
conditions for the use of DDP include sites with
alluvium, soft clays, loose sand, or chalk. DDPs
should be considered as an alternative to
conventional CFA piles in instances where a weak

21

layer is underlain by a stronger, more competent


layer(s) at moderate depths, and where potential soil
mining and effects from ground vibrations would be a
concern (Brown, 2005).
According to Bustamante and Gianeselli (1998),
the performance of DDP may be compromised due to
possible difficulties encountered during installation in
very loose sandy soils or very soft clayey soils
(characterized by SPT N-values<5 blows/0.3 m (5
blows/ft) or qc<1 MPa (145 psi)). According to Brown
(2005), the use of DDP may impractical or problematic
in predominantly saturated, fine-grained and plastic
soils where the advancement and extraction process
will cause extensive remolding of the soils, which
could be deleterious to the soil structure, shear
strength, and performance of the pile. In some
instances, displacement piles may be used even in
the presence of incompressible and/or nondisplaceable layers (e.g., dense sands and gravel,
overconsolidated cohesive layers, weathered and soft
rock) provided that (1) the thickness of the
incompressible layer is less than about 1 to 1.5 m (3
to 5 ft), and (2) the incompressible or nondisplaceable or layer is located at the lowermost strata
that is to be treated (Soilmec, 2012).
4 GENERALIZED CONSTRUCTION PROCEDURE
4.1 Advancement Phase Drilling
During the drilling phase, the tool and drill string are
rotated clockwise and penetrate the ground using the
single rotary drive and crowd force provided by the
drill rig, causing the material to move upward as the
tool moves downward. A drilling tip (appropriate for
the anticipated ground conditions) is attached to the
bottom of the drill string is used to loosen the soil
during the advancement of the tool and to prevent soil
from entering and plugging up hollow stem. As the
drill string advances deeper into the soil, the lower
auger flights move the soil upward toward the
displacement body, which then displaces and
compacts the soil radially outward into the borehole
wall and surrounding formation. The drilling phase
continues until the desired depth is achieved. The
maximum achievable depth is limited by the
capabilities of the drill rig: (a) the pull up/extraction
force, (b) the maximum available rotary torque, and
(c) the height of the drill mast, which can be extended
using a Kelly extension and/or by the addition of
jointed sections to the drill string.
4.2 Extraction Phase Concreting
Once the desired depth has been achieved, the
displacement tool and drill string are extracted while
concrete or grout is simultaneously pumped through
the hollow stem and injected into the void created by
the tool. The concrete or grout is injected either

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

22

The use of displacement piling technology in soft soil conditions

through a grout port (with disposable plug) or a


pivoting gate located near the bottom of the tool or drill
string (used with a re-usable drilling tip) or through the
bottom of the drill string (used with a sacrificial drilling
tip). The pressure used to pump the concrete or grout
should be correlated to the ground conditions at the
pile location, and should be established using a preproduction test program. NeSmith (2002) indicated
that for loose to medium-dense cohesionless soils,
the initial target lift-off grout pressure should be
within the range of about 140 to 210 kPa (20 to 30 psi)
and between 70 to 105 kPa (10 and 15 psi) for the
remainder of the grouting. As the bottom of the drill
string approaches the ground surface, the pressure
should be gradually decreased according to the in situ
stress state. When the tool is within about 1.5 to 3 m
(5 to 10 ft) of the surface, the pressure should be
decreased to zero and the pumping stopped.
As described in Section 5, rotation of the
displacement tool may or may not occur, depending
on the tool and technique utilized. The concrete mix
or the grout mix contains many similar components:
Portland cement, aggregate (fine aggregate for grout),
water, fly ash, and admixtures. The admixtures affect
and control the rate of hydration (for workability and
set time), and the water reducers (e.g., plasticizers)
affect the amount of water needed for fluidity and
flowability to ensure the fresh concrete can get to its
intended location without clogging the lines. Brown
(2005) reported that the concrete or grout mix should
provide that all solids remain in suspension without
excessive bleed-water, must be capable of being
pumped without difficulty, penetrate and fill open voids
in the adjacent soil, and allow for insertion of the steel
reinforcement.
Therefore, a high slump, fluid
concrete or grout mix (with aggregates of fine gravel
with a mximum particle size of 18 mm (-inches))
should be used.
Through the action of the
displacement tool, the borehole wall is compacted and
is relatively smooth, which reduces the concrete
overbreak and eliminates risk of over-augering.
4.3 Insertion of Steel Reinforcement
Depending upon the technique used to construct
the DDP, the steel reinforcement (cages, bars,
beams, etc.), when required, can be placed either
prior to or after the extraction of the tool and
concreting. For displacement tools that have a large
internal passage (Figure 3), it is possible to insert the
steel reinforcement inside the hollow stem prior to the
placement and injection of concrete. In this instance,
the tip at the end of the tool is sacrificial, and serves
as the injection location for the concrete. The tool is
then extracted during the concreting process.
For most techniques, however, the steel
reinforcement is inserted after the tool has been
extracted, the concreting completed, and while the
concrete is still fresh.
Depending upon depth,
reinforcement configuration, and fluidity of the

concrete, the reinforcement may need to be pushed


or vibrated into place.

Figure 3. Photograph of sacrificial tip, lower section of a


cylindrical displacement tool, and the internal passage
within the hollow stem of the tool.

5 TYPES OF DISPLACEMENT PILES & TOOLS


As Brown (2012) explained, the torque and crowd
required to construct a drilled displacement pile is
substantialthe energy required to install the pile is
related to the resistance of the soil to the
displacement, and so the piles are often installed to a
depth that is controlled by the capabilities of the
drilling rig. For the installation of conventional DDPs,
modern hydraulic drilling/piling rigs are capable of
producing high torque (500 kN-m or 370,000 ft-lb)
and large crowd forces (450 kN or 100,000 lb), which
are needed for the desired pile diameters and depths.
Paniagua (2006) provided a detailed history of the
evolution of DDPs and the principal advancements
realized during each of these three generations. Basu
et al (2010) present a comprehensive narrative and
thorough review for many of the different tools and
techniques used to construct displacement piles in
Europe and North America. It is important to note that
the advancements made in DDP technology are the
direct result of contractors and equipment/tooling
manufacturers developing practical solutions to actual
problems and issues.
The early methods (prior to the 1970s) used to
construct displacement piling (first generation piles)
focused on either soil removal during the
advancement of the tool or on the insertion of large
casing into the ground during advancement.
Moreover, relatively low torque (50-100 kN-m or
37,000 to 74,000 ft-lb) was required by the drilling
equipment to perform these piles, but the production
was slow. The methods comprising these first
generation piles include: Atlas piles, DeWaal piles,
Franki VB piles, Fundex piles, and Olivier piles. The
next version of displacement piles emerged during the
1970s and improved upon the production rate
achievable during advancement by adding partial
auger flighting near the bottom of the tooling.
Methods comprising this second generation of

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MARINUCCI A. et al.

displacement piling include: Pressodrill pile, Tubex


pile, SVB pile, and SVV pile. Since the 1980s and
even moreso during the last two decades, there has
been advancements in the tooling (e.g., increased
diameters, and design of the flights and body to
increase production), techniques (e.g., reduced
vibrations, spoils, and noise), and drilling equipment
(e.g., greater torque and pulldown crowd force, which
permit larger diameters and greater depths). Methods
comprising this third generation of DDP include:
Omega pile, Berkel Auger Pressure Grouted
Displacement (APGD) pile, Menard controlled
modulus
columns,
Trevi
Discrepiles,
and
displacement piles constructed using the Soilmec
Traction Compaction Tool (TCT).
To highlight the main differences among select
displacement piling methods, the following sections
provide a succinct overview of select techniques. A
schematic of four displacement piling methods
(DeWaal pile Fundex pile, Omega pile, and Berkel
APGD pile), photographs of different Soilmec
Discrepile tools, and a schematic of the Soilmec
Traction Compaction Tool are provided in Figures 4,
5, and 8, whereby both the displacement mechanism
can be discerned and the differences in the shape of
the completed DDP can imagined.
5.1 DeWaal Pile
During the drilling phase, the De Waal displacement
tool (Figure 4b) is advanced downward into the
formation using clockwise rotation and crowd force.
Once the desired depth is reached, concrete is placed
into the hollow stem of the drill string to a prescribed
distance above the ground surface (i.e., head), and
then the sacrificial tip at the bottom of the

23

typically inserted into the borehole after the concrete


has been placed but while it is still fresh. In some
instances, the steel reinforcement can be inserted into
the hollow stem prior to the placement of concrete.
Typical diameters achievable with this technique
range from about 300 to 600 mm (12- to 24-inches),
and to a maximum depth of about 25 m (82 ft).
5.2 Fundex Pile
For Fundex displacement piles, the tool with a conical
auger tip (Figure 4c) is advanced downward into the
formation with clockwise rotation and crowd force,
thereby displacing the soil radially outward. Once the
desired depth is reached, steel reinforcement is
inserted into the hollow stem, the sacrificial drilling tip,
which forms the enlarged pile bottom, is released, and
concrete is placed into the hollow stem. The tool/drill
string are then extracted using an oscillating up-anddown motion along with a 180 clockwisecounterclockwise rotation, while ensuring the
concrete is maintained at desired level within casing.
The withdrawal process produces a nearly smooth
shaft. The possible diameters and lengths for Fundex
piles range from about 450 to 675 mm (17.5- to 26.5inches) and to a maximum depth of about 35 m (115
ft), respectively (Basu et al, 2010).
5.3 Omega Pile
As shown in Figure 4d, the diameter of the flange
along the length of the Omega tool and partial auger
flights increases gradually and similarly from both
ends of the tool toward the displacement body (with
maximum diameter).
For displacement piles
constructed using this method, the tool (with a varying
diameter) and drill string are advanced downward into
the formation using clockwise rotation and crowd
force. Once the desired depth has been reached,
concrete is injected under pressure and the sacrificial
tip is released. During the extraction of the tool while
maintaining a clockwise rotation, concrete is injected
under pressure until the tool and drill string are fully
extracted from the borehole. The reinforcement cage
is inserted (assisted by vibratory means and/or
downward force) down into the fresh concrete. For
some Omega piles, it is possible to place the steel
reinforcement (e.g., cage or bar) into the drilling stem
prior to the placement of concrete (Bottiau, 2006).
5.4 Berkel (APGD) Pile

(a)

(b)

(c)

(d)

Figure 4. Schematic of select Displacement Pile methods:


(a) DeWaal Pile, (b) Fundex Pile, (c) Omega Pile, and (d)
Berkel APGD Pile (modified after Basu and Prezzi, 2005).

tool is released. The tool and drill string are extracted


using clockwise rotation while maintaining a constant
head of concrete within the hollow stem, resulting in a
relatively smooth surface. The steel reinforcement is

For the Berkel Auger Pressure Grouted Displacement


(APGD) pile, the tool (Figure 4e) is advanced
downward into the formation with clockwise rotation
and crowd force. Once the desired depth is reached,
high-strength grout is injected under pressure through
the hollow stem of the drill string. Once the initial
target pressure is achieved, the extraction of the tool
maintaining a clockwise rotation and grouting of the

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

24

The use of displacement piling technology in soft soil conditions

borehole is initiated.
Pressurized grouting is
continued along clockwise rotation of the tool until the
tool and drill string are fully extracted from the
borehole. After the tool is removed from the borehole
and while the grout is still fresh, the steel
reinforcement is inserted in the grouted hole.
Essentially, two types of APGD piles can be
constructed: (1) partial displacement piles in loose to
dense sands (with N<50) where the diameter ranges
from 300 to 500 mm (12- to 20-inches) and to about
17 m (80 ft) in length, and (2) full displacement piles
in loose-to-medium dense sands (corresponding to
SPT blow count N<25) where the diameter can range
from 300 to 450 mm (12- to 18-inches) and to about
24 m (79 ft) in length (NeSmith, 2002).

the downward spiraling auger flights. Eccentric-type


roller tools are able to form boreholes with diameters
ranging from about 450 to 600 mm (17.5- to 26.5inches).

5.5 Trevi Discrepile


The cylindrical (Cilindrico) displacement tool shown
in Figure 5a is well suited for soft ground conditions:
loose to medium-dense sands, soft clays, and organic
soils. The conical (Conico) displacement tool is
shown in Figure 5b, and is well suited for stronger
ground conditions: medium-dense to dense sands
and stiff clays. The conical tool is modular, and is
composed of four primary sections: (1) a drilling tip
fitted with teeth appropriate for the soil conditions
being drilled, (2) a lower section with right-handed
partial auger flights that move the soil upward toward
the displacement body, (3) a central cylindrical body
that stabilizes and displaces the soil radially outward
thereby producing the actual pile diameter, and (4) an
upper section with left-hand partial auger flights that
move the soil above the tool downward toward the
displacement body. Different pitch lengths can be
used on the auger flighting depending upon the
anticipated soil conditions: (1) a short pitch is
preferable for very loose and fine sands, soft clays,
and organic soils, (2) a medium pitch is suitable for
medium-dense sands and medium clays, and (3) a
long pitch is preferable for medium-dense to dense
cohesionless soils, medium to stiff clay, and sandy
gravel. Cylindrical and conical displacement tools are
able to form boreholes with diameters ranging from
about 350 to 600 mm (13.5- to 26.5-inches).
The eccentric roller (Pirucca) displacement tool
shown in Figure 5c is ideally suited for soft to medium
ground conditions: loose to medium-dense sands, soft
clays, and organic soils. Given the eccentric nature of
the tool, only a portion of the roller is in contact with
the soil at a time, while the remaining main portion of
the tool/drill string is away from the wall, thereby
decreasing the frictional resistance acting on the
tool/drill string. As a result, less torque is needed to
rotate and advance the tooling/drill string, which
allows the use of smaller/lighter drill rigs (lower
operational and transport costs). Soil that is not
compacted into the borehole by the eccentric roller is
moved downward toward the displacement body by

(a)

(b)

(c)
Figure 5. Different types of Soilmec Discrepile tooling:
(a) cylindrical tool, (b) conical tool, and (c) eccentric
roller tool.
The requirements for proper selection of a drill rig
capable of constructing displacement piles using the
Discrepile tools include (a) a rotary head capable of
delivering rotation at about 20 to 25 rpm, (b) a rotary
head capable of delivering at least 200 to 250 kN-m
(147,000 to 185,000 ft-lbs) of torque, (c) a pull down
system with a crowd force of at least 200 kN (45,000
lb); and (d) a pull-up system capable of providing an
extraction force of at least 200 kN (45,000 lbs). The
concrete pump should be sized according to the
expected extraction rate of the drill string, the volume
of the void created by the displacement tool as it is
extracted, and the required pressure range that will be
used during the injection of the concrete.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MARINUCCI A. et al.

5.5.1 Case History La Prua Business Center;


Rimini, Italy
To support the proposed mixed-use residential and
business structures for the new La Prua Business
Center located along the waterfront in Rimini (Figures
6 and 7), the design engineer estimated that more
than 20,000 linear meters (65,000 linear feet) of bored
and/or driven piling would be required to support the
building structures.
At this site, the general
subsurface profile consisted of a layer of silt with sand
and clay to a depth of about 3.1 m (10 ft) underlain by
a 7.8 m (25.5 ft) thick layer of medium-coarse sand
with silt and rounded pebbles, which was underlain by
a 14.05 m (46 ft) thick layer of clayey silt with traces
of organics. Beneath the clayey silt layer, there is a
2.1 m (6.9 ft) thick layer of medium sand with silty
gravel, underlain by a 1.1 m (3.6 ft) layer of gravel with
sand and silt, which is then underlain by a 6.8 m (22.3
ft) thick layer of clayey silt and clay with interbedded
lenses of sand.

25

The original foundation options (Figure 8) included


bored piles stabilized with bentonite slurry during
drilling and driven piling. However, it was deemed that
the driven piling operations would have caused
excessive environmental concerns (e.g., noise and
vibrations) to the nearby residents and surrounding
buildings, respectively. In addition, there was concern
that the bored pile operations, especially the use of
bentonite slurry, would have issues related to the
cleanliness of the jobsite and effect on the
surrounding roads resulting from the trucks
transporting the excavated and removed the spoils.

Figure 8. A portion of the plan view of the foundation


structure layout beneath the curved structure and the middle
structure.

Figure 6. Aerial photograph of the location (outlined in red)


for the proposed mixed use, residential-commercial
structure.

As an alternative to conventional piling, the


geotechnical specialty contractor, Trevi S.p.A.,
proposed using unreinforced displacement piles
installed in a diamond shaped pattern for ground
improvement beneath the structures (Figure 9).
There was concern expressed by the owner / engineer
that the unreinforced displacement pile elements
would not provide structural support should it be
required or needed. To ensure adequate support for
the structures and allay any concern by the owner, the
contractor performed a compression test on a
sacrificial, non-production displacement pile (Figure
10). As shown in Figure 11, at a load of about 176
tons, the top and creep displacements were
approximately 8.5mm (5/16 inch) and 3mm (1/8 inch),
respectively.

Figure 7. Axonometric illustration of the proposed mixed


use, residential-commercial structure.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

26

The use of displacement piling technology in soft soil conditions

(a)
(a)

(b)
Figure 9. (a) Trucated plan view showing the pattern of the
installed displacement piles, and (b) photograph of two
exposed tops of completed displacement piles.

(b)
Figure 11. Graphical depiction of the load-displacement
behavior from (a) a compression load test on an
unreinforced displacement pile, and (b) a creep load test
performed with a constant load maintained for 60 minutes.

In total, about 1,600 piles with a diameter of about


600 mm (24-inches) and an approximate length of 25
m (82 ft) were installed using a conical displacement
tool and a Soilmec SR-65 drill rig. Due to the
compactive nature of the tool and the resulting face of
the borehole wall, the concrete overbreak was kept to
a minimum, as anticipated, and averaged between
5% and 10% above theoretical. In addition, the drill
spoils that needed to be disposed were also kept to a
minimum, where the disposal volume averaged
between 5% and 10% of the total volume of installed
piling.
5.5.2 Case History Monselice Hospital; Monselice,
Italy
Figure 10. A photograph of the compression test load
frame.
Note: a steel sleeve was used to laterally
restrain/support the exposed portion of the displacement
pile.

Located southeast of the Euganean Hills and


southwest of Padua in Italy, the Monselice Hospital is
located in a town with a population of 18,000
inhabitants and in an area with substantial geothermic
activity, leading to the popularity of local hot springs

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MARINUCCI A. et al.

and mud spas. By working with nature, the buildings


designers incorporated the use of the underground
geothermal resource to heat and cool the hospital.
The foundation for the hospital was constructed using
displacement piling technology to maximize the
benefit of minimal removal of the drilling spoils, noise,
and vibration. The geology below the hospital was
very amenable to displacement piling construction
because it was characterized by fine to mediumdensity sand (SPT N-values45) and silty clay with the
groundwater table located about 0.9 m (3 feet) below
the ground surface.
The displacement piles were installed by Trevi
S.p.A. using a Soilmec SR-80 drill rig fitted with the
displacement pile kit and the conical displacement
tool (Figure 12). The in situ soil was displaced and
compacted radially during downward penetration
during the drilling phase.

Figure 12. Installation of displacement piles using a SR-80


drill rig and a conical displacement tool.

During the extraction phase, concrete was injected


at the tip of the tool as the drill string was extracted.
When the tool was out of the hole, the reinforcement
cages equipped with geothermal loops were then
installed in the fresh concrete. The geothermal loops
were incorporated into the displacement piles to take
advantage of the stable subterranean temperatures
where to provide the hospital with heat in the winter
and air conditioning in the summer. The piles were
about 600 mm (24-inches) in diameter and ranged in
length from about 17 to 24 m (55 to 80 ft), resulting in
a total of about 44,000 linear meters (144,300 linear
feet) of displacement piling. The average daily
production of displacement piling was about 16 piles
per day or about 350-380 m per day (1,150 to 1,250
ft/day).
To provide control and monitoring during the drilling
and concreting phases, the contractor utilized the
monitoring system (Soilmec Drilling Mate System,
DMS) that was integrated into the SR-80 drill rig.

27

The monitoring system allowed the drill rig operators


and jobsite personnel to monitor and accurately
control the machine (e.g., drilling parameters and rig
performance) in real time. Data from the DMS was
also streamed to a remote computer where jobsite
managers were able to monitor, process, and
document the project information.
6 SOILMEC TCT A NEW DDP TOOL
For conventional displacement piling, as described
above, where the soil is compacted during the drilling
or penetration phase, a large, heavy, powerful drill rig
is required to provide the crowd force and torque
needed to achieve the desired displacement,
diameter and depth. Therefore, an effective manner
to reduce the need for very large drill rigs for this work
is to reduce the required crowd force and torque
needed to push the drill string and tooling into the
ground. Soilmec S.p.A. patented a displacement
piling technique using the Traction Compaction Tool
(TCT) where the soil is compacted during the
extraction phase of the pile construction instead of
during the drilling phase, thereby reducing the amount
of torque and crowd force required to turn the tooling
and penetrate the ground.
As shown in Figure 13, the TCT is composed of
three main parts: (1) the lower section contains a
drilling tip fitted with appropriate teeth (based on soil
type) to facilitate penetration into the ground, a flight
auger, a short drill string, and a concrete pivot gate
that is connected to the hollow stem of the drill string;
(2) the middle section contains flights and borehole
stabilizer that is allowed to partially freely rotate
around the hollow stem of the drill string, and are used
to displace and compact the soil during extraction and
concreting; and (3) the upper section is fitted with
flights (only a few) rigidly connected to the drill string,
which facilitate movement of the soil toward the
displacement body during extraction. The upper and
lower sections of the TCT are fixed and turn
simultaneously. The special shape of the lower tip
creates a separation with respect to the flight of the
central tools portion, and the mechanical gate
separates the tip section (where the concrete flows
out from the hollow stem) from the main section of the
tool.
By maximizing the drill equipment operability with
the TCT, comparatively smaller rigs can be used to
achieve similar sized (diameter and depth)
displacement piles constructed using conventional
displacement piling tools/drill rigs, resulting in reduced
operating and transport costs without sacrificing
quality and productivity. In addition to the benefits
achievable with conventional DP methods, the
advantages of constructing DPs using the TCT
include the construction of larger diameter elements
(up to 800 mm (32-inches)), the use of smaller/lighter

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

28

(a)

The use of displacement piling technology in soft soil conditions

(b)

Figure 13. Execution phases for Soilmec Traction


Compaction Tool (TCT) (a) drilling phase minor and (b)
extraction phase - soil is displaced and compacted radially
into the surrounding soil.

drilling rigs, and higher quality of finished element


because the mechanical gate separates the soil being
displaced from the concrete being injected.
During the development and trial phases of
development, it was determined that displacement
piles with diameters ranging from about 400 to 800
mm (16- to 32-inches) could be constructed using the
TCT. Moreover, similar to conventional displacement
piling and depending on the capability of the drill rig,
the maximum achievable depth using the TCT is
approximately 32.5 m (106 ft). (Note: at this time, the
TCT has not been utilized in large-scale, production
work; consequently, comparative case histories, load
tests, and other performance data are not yet publicly
available.)
Similar to conventional displacement pile
construction, there are two distinct phases utilized
during the construction of TCT displacement piling.
First, during the drilling phase, the TCT (in open
position, Figure 14a) is advanced into the ground and
the disturbed in situ soil fills the flight of the tool but
there is little-to-no compaction occurring during this
phase. Second, during the concreting phase, the TCT
(closed position, Figure 14b) is extracted and the soil
along the length of the pile is displaced radially into
the borehole wall.
During the drilling phase, the tool and drill string are
rotated in a clockwise rotation and penetrate the
ground using the single rotary drive and crowd force
provided by the drill rig, causing the material to move
upward as the tool moves downward. Compared to
the length of a typical continuous flight auger, the
length of the TCT is short and the flights are few,
which reduces the total friction that develops between
the soil and the tool (on the flights and hollow stem).
As such, the torque and crowd force needed by the
drilling rig to turn and advance the tool are also
reduced.

Once the desired depth has been achieved, the


rotary drive turns the drill string in a counterclockwise
direction; however, since the middle portion of the tool
(i.e., middle flights and stabilizer in Figure 13a) is in
intimate contact with the soil, the friction developed
between the stabilizer and the ground does not allow
this portion of the tool to rotate. In addition, as the drill
string is rotated counter-clockwise, the lowermost
flight (located below the stabilizer during the drilling
phase) rotates into a position that essentially forms a
cover plate above the concrete gate, effectively
separating the soil to be displaced with the area to be
concreted during extraction. During extraction, the
soil above the cover plate along the tool is rotated
downward toward the stabilizer, which is then
compacted radially outwards and into the borehole
wall while concrete is pumped through the hollow drill
string and out the pivot gate. Through the action of
the displacement tool, the borehole wall is compacted
and is relatively smooth, which reduces the concrete
consumption (i.e., overbreak) and eliminates risk of
over-augering. If required, steel reinforcement is
placed after the tool is extracted from the hole and
while the concrete is still fresh.

(a)

(b)
Figure 14. Soilmec R-625/SR-65 rig fit with the TCT
advanced DP tooling: (a) tool in open position for the drilling
phase, and (b) in the closed position that is used for the
extraction/concreting phase

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MARINUCCI A. et al.

7 QUALITY CONTROL
During construction, it is essential to control and
monitor the various parameters that affect the integrity
and performance of a displacement pile. During the
drilling phase, the important parameters include the
drilling depth, penetration speed, rotation speed of the
tool, inclination of the tool guide mast, rotary torque,
and crowd force. During the extraction and concreting
phase, the important parameters include the depth,
lifting speed, rotation speed of the tool, inclination of
the tool guide mast, rotary torque, extraction force,
concrete pressure and flow, total volume of pumped
concrete, and concrete overbreak. During extraction,
care must be exercised to coordinate the concrete
pumping rate with the extraction rate of the tooling/drill
string. Necking and other integrity issues may occur
if the tooling/drill string is extracted too quickly
comparatively, which would be realized in a pressure
decrease on the gauge at the concrete pump. The
different parameters can be continuously monitored
and recorded using automated monitoring systems
that are integrated directly into the drilling rig, which
are common on modern displacement drill rigs.
Controlling and monitoring the various parameters
during drilling and concreting assists with ensuring the
quality of the finished product consistently meets
project specifications.
8 CONCLUSION
The use of drilled displacement piles has increased
significantly during the past two decades as a result of
various factors including advancements in tooling
(e.g., increased diameters, increase production rates)
and equipment capabilities (e.g., greater torque and
pulldown force permitting larger diameters and
greater depths). The various benefits resulting from
the use of environmentally friendly displacement piles
were presented, and include minimal drill spoils,
reduced ground vibrations, and an increase in unit
side friction and end bearing resistance. As long as
the soil can be displaced and compacted, the
technique is ideally suited for a wide range of ground
conditions ranging from soft-to-firm ground conditions
and from sandy gravel to clay. As such, displacement
piling has been used for a wide array of applications
on both public- and commercial-type projects ranging
from unreinforced columnar elements for ground
improvement purposes (e.g., column-supported
embankments) and as structural foundation elements
(e.g., support for column loading). For conventional
displacement piles, the maximum diameter and depth
that can be achieved range from about 300 to 800 mm
(12- to 32-inches) and from about 24 to 35 m (80 to
115 ft), respectively, which are directly correlated with
the capabilities of the drilling rig. This paper provided
an overview of the evolution of displacement piles,
various types of techniques and tools, applicable

29

ground conditions where the technology is suitable,


and general requirements for the construction of
displacement piles.
Recent advancements to
displacement piling (e.g., Traction Compaction Tool)
were presented, where piles can be constructed to a
maximum diameter and depth of 800 mm (32-inch)
and 32.5 m (106 ft), respectively, which also facilitates
the use of comparatively smaller/lighter drill rigs, and
lower operating and transport costs.
REFERENCES
Basu, P., Prezzi, M., and Basu, D. (2010). Drilled
Displacement Piles Current Practice and Design.
DFI Journal, v.4 n. 1, August. P. 3-20.
Baxter, D.J., Dixon, N., Fleming, P.R., and Hadley,
S.P. (2006). The Design and Formation of Bored
Displacement Piles A United Kingdom
Perspective. Proceedings of the DFI/EFFC 10th
International Conference on Piling and Deep
Foundations, Amsterdam, Netherlands.
Bottiau, M. (2006). Recent evolutions in deep
foundation technologies. Proceedings of the
DFI/EFFC 10th International Conference on Piling
and Deep Foundations, Amsterdam, Netherlands.
Brown, D.A. (2005). Practical considerations in the
selection and use of continuous flight auger and
drilled displacement piles. Advances in Designing
and Testing Deep Foundations. Geotechnical
Special Publication No. 132, ASCE, pp. 1-11.
Brown, D.A. (2012).
Recent Advances in the
Selection and Use of Drilled Foundations.
Proceedings of the GeoCongress 2012: State of
the Art and Practice in Geotechnical Engineering.
Geotechnical Special Publication No. 225.
Sponsored by Geo-Institute of the ASCE. Hryciw,
R.D., Athanasopoulos-Zekkos, A., and Yesiller, N.,
editors.
Brown, D.A. (2005).
Selection and Design of
Continuous Flight Auger and Drilled Displacement
Piles. Proceedings of the Geo-Frontiers Congress
2005. Geotechnical Special Publication No. 132 of
the ASCE. Rathje, E.M., editor. pp. 1-11.
Bustmante, M. and Gianeselli, L. (1998). Installation
parameters and capacity of screwed piles. Deep
Foundations on Bored and Auger Piles, BAP III.
Balkema, Rotterdam. pp. 95-108.
Chiarabelli, M. (2015). Soilmec Introduces New,
Advanced Technologies for Displacement Piling.
ADSC-IAFD Foundation Drilling Magazine.
February/March, p. 34-37.
NeSmith, W.M. (2002). Design and installation of
pressure-grouted
displacement
piles.
Proceedings of the Ninth International Conference
on Piling and Deep Foundations, Nice, France, pp
561-567.
NeSmith, W.M. (2004). Application of Augered,
Cast-in-Place Displacement (ACIPD) Piles in New
York City. Proceedings of the Deep Foundation

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

30

The use of displacement piling technology in soft soil conditions

Institute (DFI) Augered Cast-in-Place Piles


Committee Specialty Seminar, McGraw-Hill
Building, New York, NY, pp 7783.
NeSmith, W.M. and Fox, J. (2009). Practical
Considerations for Design and Installation of Drilled
Displacement Piles.
Proceedings of the
Contemporary Topics in Deep Foundations:
Selected Papers from the 2009 International
Foundation Congress and Equipment Exposition.
Geotechnical Special Publication No. 185 of the
ASCE. Iskander, M., Laefer, D.F., and Hussein,
M.H., editors. pp. 438 446.
Pagliacci, F. (ed). (2015). Ground Engineering
Technologies. The Trevi Group. pp. 328.
Pagliacci, F. and Chiarabelli, M. (2015). Monselice
Hospital, Displacement Piling Case History. Pile
Buck Magazine, v.31, n. 2 P. 14-17.
Paniagua, W.I. (2006). Construction of Drilled
Displacement and Auger Cast in Place Piles.
Proceedings of the International Symposium: Rigid
Inclusions in Difficult Soft Soil Conditions, TC36,
Mxico DF
Prezzi, M. and Basu, P. (2005). Overview of
construction and design of auger cast-in-place and
drilled Displacement piles. Proceedings of the DFI
30th Annual Conference on Deep Foundations,
Chicago, U.S.A., pp. 497-512.
Siegel, T.C., NeSmith, W.M., NeSmith, W.M., and
Cargill, P.E. (2007). Ground improvement resulting
from installation of drilled displacement piles.
Proceedings of the DFI 32nd Annual Conference on
Deep Foundations, Colorado Springs, U.S.A., pp.
129-138.
Soilmec S.p.A. (2013). DMS Drilling Mate System.
pp. 28.
Soilmec S.p.A. (2012). Displacement Piles
Technology. pp. 16.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Rescate de una cimentacin de pilas con inclusiones rgidas


Pile foundation retrofit with rigid inclusions
Jos SEGOVIA1, Walter PANIAGUA2 & Germn LPEZ3
1Pilotec

S.A. de C.V., Mxico


S.A. de C.V., Mxico
3Facultad de Ingeniera, UNAM, Mxico
2Pilotec

RESUMEN: La construccin de una torre de consultorios mdicos en el sur de la zona de transicin de la ciudad de Mxico,
entre el Pedregal de San ngel y el Cerro de la Estrella, motiv al diseador geotcnico a resolver la cimentacin con
pilas coladas in situ. Sin embargo, debido a que la construccin de pilas la realiz una empresa con escasa experiencia
en este tipo de trabajo, la supervisin de la obra solicit la evalucin de las pilas mediante pruebas de integridad. El
resultado de esta revisin llev a la conclusin de redisear la cimentacin mediante una losa, utilizando las pilas solamente
como inclusiones rgidas para reducir el asentamiento del edificio.

1.2 Ubicacin geotcnica y solucin de cimentacin

1 ANTECEDENTES
1.1 Caractersticas del Proyecto
El proyecto consiste en la construccin de la
cimentacin de dos torres de 6 niveles y azotea, las
cuales se construirn en etapas. En la Figura 1 se
muestra una planta con la localizacin general.

Figura 1. Localizacin de Hospital

El sitio se ubica dentro de la zona de Transicin que


se caracteriza por un costra superficial, subyaciendo
por arcillas limosas de alta plasticidad y depsitos
fluvio lacustres y/o tobas areno limosas de alta
resistencia al corte y baja deformabilidad, Figura 2.

Figura 2. Ubicacin del proyecto en la zona de Transicin

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

32

Pile foundation retrofit with rigid inclusions

En la Fig. 3 se muestra una planta general del


proyecto, con la ubicacin de los sondeos de
exploracin realizados: dos sondeos mixtos, dos
sondeo de muestreo selectivo y un tubo de
observacin del NAF.

De acuerdo con proyecto, la solucin de la


cimentacin para el edificio fue con base en pilas
coladas en el lugar, con una profundidad de 19.0m
respecto al nivel de banqueta, apoyadas en los
estratos resistentes indicados por el estudio de
mecnica de suelos como fluvio lacustres con
dimetros de 0.8, 1.0 y 1.20 m. Las pilas de
cimentacin se complementaron con una losa tipo
waffle con trabes de 0.80 m de peralte y dados de
1.6 m de profundidad para unir las pilas, Figura 5.
Debido a diferentes circunstancias, la supervisin
de la obra solicit la revisin de la integridad
estructural de las pilas y de su longitud de desplante,
para lo que se solicit la realizacin de una serie de
pruebas de integridad (PIT) para determinar la
sanidad de la pilas.

Figura 3. Planta general del proyecto

Los suelos encontrados de 0 a 5.4m, comprenden


limos arcillosos y arenosos, de consistencia media a
alta. De 5.4m a 18m, son suelos arcillosos y limosos
de alta plasticidad, baja resistencia y alta
compresibilidad, de consistencia blanda a media.
Finalmente, de 18m y hasta 30m (mxima
profundidad explorada) hay depsitos fluviolacustres, formados por arenas finas y limos
arenosos, muy compactos y de consistencia muy
rgida, Figura 4. El NAF se detect a 2.15m de
profundidad.

Figura 5. Planta de pilas de cimentacin del Hospital

2 PRUEBAS DE INTEGRIDAD
2.1 Generalidades
La prueba de integridad es un ensayo para de
determinar la variacin de las caractersticas del
concreto de las pilas de cimentacin en toda su
longitud. La forma usual del ensayo consiste en la
colocacin de un acelermetro en el cabezal de la pila
bajo prueba, y en la aplicacin de golpes con un
martillo instrumentado Figura 6.
2.2 Descripcin de la prueba

Figura 4. Caracterizacin geotcnica y diagrama de


esfuerzos

El acelermetro se fija a la cabeza de la pila por medio


de cera de petrleo. Los golpes del martillo generan
una onda de compresin, que recorre la pila y sufre
reflexiones al encontrar cualquier variacin en las
caractersticas del material. Esas reflexiones causan
variaciones en la aceleracin medida por el sensor. El
equipo hace un registro de la evolucin de esa
aceleracin con el tiempo.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

SEGOVIA J. et al.

Figura 6. Equipo de medicin para pruebas de integridad

Es usual la aplicacin de varios golpes


secuenciales a fin de que el equipo de la prueba de
integridad obtenga un promedio de las seales
correspondientes. Ello permite eliminar interferencias
aleatorias o efectos anmalos, sobresaliendo en la
seal las variaciones causadas por las reflexiones de
la onda. Como la onda hace su recorrido con una
velocidad fija (velocidad del concreto), al conocerse
esa velocidad de propagacin y el tiempo transcurrido
entre la aplicacin del golpe y la llegada de la reflexin
correspondiente a las anomalas o a la punta, es
posible determinar la localizacin exacta de stas.
Las vibraciones superficiales son grabadas en la
parte superior del cabezal de la pila as como todas
las reflexiones primarias. Considerando la naturaleza
y los tiempos de observacin de las reflexiones, es
posible valorar la integridad de la pila y detectar
anomalas.
El reflector ms profundo es la punta de la pila
(parte baja), por lo que su reflexin es la ltima que
puede observarse, Figura 7.

33

y su longitud, L, la reflexin de la pila es esperada al


tiempo 2L/C.
Las reflexiones se observan mediante un
acelermetro mvil, temporalmente adosado en el
cabezal de la pila. La aceleracin de la seal es
digitalizada y almacenada en el equipo colector.
Asimismo, la aceleracin se integra numricamente
para producir un velocigrama. Por otra parte,
midiendo la aceleracin del martillo y al multiplicarla
por su masa, es posible conocer el valor de la fuerza
aplicada, F. Durante el impacto, cuando el martillo y
la pila estn en contacto, la fuerza es proporcional a
la velocidad esperada del cabezal de la pila. La
constante de proporcionalidad es la impedancia
acstica Z en el cabezal.
El cociente F/Z es llamado velocidad y es
presentado conjuntamente con la velocidad
observada V. Ambas deben de tender a ser iguales
durante el impacto. La fuerza grabada despus del
impacto no tiene an un significado prctico desde el
punto de vista de la integridad de la pila.
Para facilitar la localizacin de los reflectores, la
seal se presenta en funcin de la distancia mediada
desde el punto de impacto.
Las reflexiones son generadas por las variaciones
de la impedancia de la pila, Z, que puede calcularse
con,
=

(1)

donde A = rea transversal de la pila; densidad del


material que conforma la pila; C = producto que
indica cambio de impedancia y que genera
reflexiones.
Con la geometra de la pila (rea transversal y
longitud), es posible detectar irregularidades en la
pila, tanto la variacin de la seccin transversal a lo
largo del fuste como los cambios en el tipo de
material,
mediante
la
interpretacin
del
comportamiento anmalo observado entre la seal
incidente y el reflector de la punta, Figura 7.
2.4 Interpretacin de resultados
Se estudiaron el 100% de las pilas. En las Figuras 8
y 9, se muestran algunas de las pilas ensayadas.
La interpretacin de los registros de campo se
realiz con la ayuda de los programas de cmputo
PIT-W-2003 y Profile-2003, donde es posible realizar
el post-procesado de los registros tanto en el dominio
del tiempo como en el de la frecuencia.

Figura 7. Esquema de la reflexin de ondas en caso de


disminucin de seccin de una pila

2.3 Interpretacin de la prueba


Dada una estimacin de la velocidad de onda que
viaja por el material del que est compuesta la pila, C,

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

34

Pile foundation retrofit with rigid inclusions

Figura 8. Pilas ensayadas

Figura 9. Pilas ensayadas

2.5 Resultados
La interpretacin de los registros de campo se realiz
con la ayuda de los programas de cmputo PIT-W2003 y Profile-2003, donde es posible realizar el postprocesado de los registros tanto en el dominio del
tiempo como en el de la frecuencia.
Es importante mencionar que la prueba de
integridad da solamente una calificacin cualitativa de
la pila, por lo que no debe utilizarse como un
mecanismo de clasificacin de la pila y de su proceso
constructivo. Es importante indicar que las
ampliaciones y rugosidades en las pilas no deben
considerarse como un defecto, sino en la mayora de
los casos como un incremento en su capacidad de

carga tanto por friccin como por punta, con la


condicin de que exista continuidad en el elemento.
La calificacin de la calidad de la pila se agrupa en
familias de calidad:
A. Pila Buena: No se aprecian defectos obvios y la
respuesta de la punta de la pila es clara para
longitudes de pila de hasta 30 dimetros.
B. Pila Mala: Existe una identificacin clara de
defectos en la pila, no se aprecia claramente el
reflector de la punta de la pila despus de eliminar el
ruido
de
la
seal,
aun
cuando
se
cumple con el criterio de que la longitud de la pila es
menor que 30 veces su dimetro.
En tales casos, para poder descartar una pila es
recomendable llevar a cabo pruebas de detalle
(pruebas de carga y/o sondeos de inspeccin),
analizar profundamente el historial de construccin de
la pila, as como realizar correcciones en el caso de
que los defectos se muestren superficialmente y
volver a efectuar los ensayes de integridad.
C. Pila con posibles defectos. Los defectos en la
pila no son claros. Es necesario llevar a cabo pruebas
de integridad adicionales despus de aplicar medidas
correctivas en caso de que los defectos se localicen
en la parte superior (gran longitud del armado que
sobresale en el cabezal, imperfecciones del cabezal,
deficiencia en el pulido de la superficie del cabezal en
la zona donde se coloca el acelermetro, etc.); en
caso que los posibles defectos persistan, ser
necesario llevar a cabo pruebas de carga o sondeos
directos (extraccin de ncleos) para poder descartar
dicha pila.
D. Datos no concluyentes. No se tienen registros
de calidad debido a imperfecciones en el cabezal de
la pila (armado que sobresale, superficies mal
pulidas, contaminaciones en el cabezal), a la alta
resistencia del terreno localizado a lo largo de la pila
o debido a la longitud de la pila por lo que el pulso
reflector de la punta no sea observable (un criterio
emprico para definir la longitud mxima de la pila
para que pueda observarse dicho pulso reflector es
que la longitud de la pila sea menor que 30 veces su
dimetro).
Velocigramas e interpretacin de resultados
Los velocigramas capturados en campo, fueron
interpretados en gabinete mostrando los siguientes
resultados.
Se probaron la totalidad de las pilas de la
estructura, 49 pilas; 22 pilas califican sin defectos, de
acuerdo a las pruebas; 8 pilas muestran reduccin en
su dimetro o menor longitud que la de proyecto y 19
pilas se manifiestan con posibles defectos
estructurales, Figura 10.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

SEGOVIA J. et al.

35

DN P31-6

6: # 7
-0.15

0.00

0.15

0.30

cm/s
0

Wavelet
Impedance:
Mass:

2.92 m

10

10

12

15

14
15.04 m (4200 m/s)

Relative Vol.:
Construct. Vol.:
Construct. Area:
Max Prof ile:
Min Profile:

0.74
1.00
1.00
1.39 at 2.04 m
0.40 at 13.98 m

x 39
16

20

18

20

25

22
Magn

V 0.210 cm/s (0.350)


F/Z 0.013 cm/s (0.016)

diam

Figura 12. PILA MALA. Se observan decrementos del


dimetro de la pila a partir de los 8m de profundidad, as
como un decremento de la longitud de la pila de 15.15 a
14m
DN P30-6

6: # 7
-0.25

0.00

0.25

0.50

cm/s
0

Figura 10. Resultado de la interpretacin de las pilas

Low Pass:
Wavelet
Impedance:
Mass:

0.26 m
1.78 m

10

10

12

15

14
15.04 m (4200 m/s)

En las Figuras 11, 12 13 y 14, se presentan


ejemplos de los velocigramas de campo y su
interpretacin de gabinete.

Relative Vol.:
Construct. Vol.:
Construct. Area:
Max Prof ile:
Min Profile:

0.91
1.00
1.00
1.37 at 1.91 m
0.74 at 13.45 m

x 20
16

20

18

20

25

22
Magn

0.00

diam

Figura 13. PILA CON POSIBLES DEFECTOS, Se observan


decrementos del dimetro a partir de los 12m de
profundidad hasta la punta de la pila, as mismo se observa
un decremento de la longitud de la pila de 15.45 a 15m. Se
aprecia claramente la seal de la pila

DN P28-5

5: # 4
-0.40

V 0.173 cm/s (0.210)


F/Z 0.012 cm/s (0.013)

0.40

0.80

cm/s
0

Low Pass:
Wavelet
Impedance:
Mass:

0.36 m
0.42 m

5833 Hz
5000 Hz
1.143E+004 kN/m/s
0.9 kg

5
DN P34-6

4: # 8-175%
-2.75

0.00

2.75

5.50

cm/s

12

16.11 m (4200 m/s)


x 20

10

Relative Vol.:
Construct. Vol.:
Construct. Area:
Max Prof ile:
Min Profile:

1.09
1.00
1.00
1.66 at 1.73 m
1.00 at 0.00 m

Low Pass:
Wavelet
Pivot
Impedance:
Mass:

1.76 m
1.56 m

1.143E+

16
6

15
8

20
10

12

Magn

V 0.365 cm/s (0.388)


F/Z 0.009 cm/s (0.009)

24
m

20
diam

10

14
15.15 m (4200 m/s)
x 20

16

Figura 11. PILA BUENA. Se observa un incremento


importante de la longitud de la pila de 15.15 a 16.1m

18

15

20

22
Magn

V 0.929 cm/s (1.486)


F/Z 0.005 cm/s (0.006)

diam

Figura 14. DATOS NO CONCLUYENTES. No se observa


una seal clara

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Relative Vol.:
Construct. Vol.:
Construct. Area:
Max Prof ile:
Min Profile:

0.70
1.00
1.00
1.38 at 1.79 m
0.20 at 11.21 m

36

Pile foundation retrofit with rigid inclusions

3 ALTERNATIVA DE SOLUCIN

Carga de la estructura

En las condiciones actuales las pilas de cimentacin


no se consideran aptas para trasmitir las cargas de la
estructura a los depsitos resistentes del subsuelo de
la zona; sin embargo, esta situacin puede ser
solventada utilizando las pilas como inclusiones
rgidas que restrinjan la deformabilidad de los
depsitos lacustres que se localizan entre 5.4 y 18.0
m de profundidad.

Loas de cimentacin y plataforma


de transferencia

3.1 Antecedentes de las inclusiones rgidas


Las inclusiones rgidas han sido empleadas desde
hace muchos aos como un sistema para reducir
asentamientos diferenciales e incrementar la
capacidad de carga de estratos blandos; se han
utilizado principalmente como solucin de grandes
plataformas de almacenamiento en Europa y Oriente,
pero grandes ejemplos utilizando estacones de
madera como inclusiones rgidas lo tienen muchos
edificios del centro histrico de la Cd. de Mxico.
En los aos recientes grandes desarrollos
habitacionales han sido construidos utilizando este
principio, principalmente en la zona de Aragn y en
ciudades como Morelia, Michoacn y como ejemplo
notorio de su utilizacin se tiene el rescate de la
Catedral Metropolitana de la Cd. de Mxico.

Inclusiones rgidas

Figura 15. Caractersticas de una cimentacin con


inclusiones

Figura 16 Esquema de solucin

3.2 Caractersticas de la cimentacin


Una de las claras ventajas que se tienen al utilizar las
pilas como inclusiones rgidas, es que se aprovechan
las pilas que se encuentran construidas, modificando
su concepto de trasmisoras de carga a reductoras de
deformacin.
Bajo el concepto anterior se requiere que la losa de
cimentacin tenga la suficiente rigidez para trasmitir
de una manera uniforme las cargas al subsuelo,
adems de dotar a la cimentacin de una plataforma
de transferencia entre la losa y las pilas de
cimentacin, Figura 15.
La plataforma de transferencia est compuesta por
una capa de tepetate compactado, suelo cemento o
relleno fluido de baja resistencia.
Otro punto importante de la solucin es que se
mantienen los niveles arquitectnicos sin modificar
ninguna parte del proyecto.
3.3 Diseo conceptual
En relacin con la magnitud de cargas que se
trasmitirn a la solucin de inclusiones se tiene que:
La estructura pesa alrededor de 5,400 t
rea de cimentacin 995 m2
Descarga neta 5.4 t/m2

4 PROCEDIMIENTO CONSTRUCTIVO
La implementacin de la solucin requieri de las
siguientes acciones:
Demoler las cabezas de las pilas hasta el nivel 3.95 m respecto al 0.0 de proyecto.
Colocar como capa de transferencia un limo
arenoso (tepetate) compactado al 95% de la prueba
Prctor estndar, colocado en capas mximas de 20
cm y con un adecuado control del contenido ptimo
de agua de compactacin; la capa se coloc desde el
nivel de demolicin (-3.95 m) hasta el nivel -3.25 m.
Construccin de las contratrabes de cimentacin,
se estim un peralte de 80cm, cuyo valor final se
derivar
de
los
anlisis
estructurales
complementarios.
Colocacin entre contratrabes de una capa de
tepetate compactado al 90% Prctor.
Construccin de la losa de cimentacin de acuerdo
con los resultados de los anlisis estructurales.
Los peraltes de las contratrabes y de losa de
cimentacin dependern de los resultados finales de
los anlisis estructurales.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

SEGOVIA J. et al.

5 CONCLUSIONES
Ante una mala prctica de construccin de pilas, el
proyecto de la torre de consultorios se vio
comprometido.
La solucin de desligar las pilas de la losa de
cimentacin y analizarla como una cimentacin con
base en una losa de concreto, reforzando el suelo con
inclusiones, rehabilit la viabilidad del proyecto y
permiti concluir su construccin, Figura 16.

Figura 16 Estado de la estructura, octubre 2015

REFERENCIAS
SMMS (2002), Manual de Construccin Geotcnica,
Cap. 9, Inclusiones, Sociedad Mexicana de
Mecnica de Suelos.
Santoyo, E. y Ovando, E. (2001), Catedral y Sagrario
de la Ciudad de Mxico, Correccin Geomtrica y
Endurecimiento del Subsuelo, editado por TGC
Geotecnia S.A. de C.V.
Paniagua, W.I., Ibarra, E. and Valle, J.A. (2008),
Rigid Inclusions for Soil Improvement in a 76
Building Complex, 33rd Annual Members
Conference, New York, Deep Foundation Institute.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

37

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Foundation design and construction for high-rise Towers in Mexico City


Diseo de la cimentacin y construccin de Torres de gran altura en la Ciudad de Mxico
Peter W. DEMING1, Sissy NIKOLAOU2, Raymond J. POLETTO3, & George J. TAMARO3
1Mueser

Rutledge Consulting Engineers, Partner, New York, NY USA


Rutledge Consulting Engineers, Senior Associate, New York, NY USA
3Mueser Rutledge Consulting Engineers, Consultant, New York, NY USA

2Mueser

ABSTRACT: The unique Mexico City subsurface conditions with deep rock and natural valley topography filled with soft
plastic clays, combined with the high seismic activity, makes design and construction of foundations for high-rise
buildings challenging. The authors present some of the challenges stemming from more than two decades of relevant
experience that include: (i) need for comprehensive site characterization including in-situ dynamic measurements; (ii)
seismic hazard, ground motions that reflect the amplification of the seismic waves propagating from rock through deep
soft clay; (iii) high risk that temporary excavations will cause ground movement and damage adjacent structures (iv)
settlement due to the clay behavior and with respect to regional settlement; (v) soil-foundation-structure interaction; and
(vi) dense urban construction that could result in building-to-building interaction. Recent projects using a performancebased design approach are discussed, raising some questions and presenting some solutions. The paper provides an
overview of foundation systems used for few high-rise buildings of 40 or more stories along Paseo de la Reforma.

1 INTRODUCTION
Foundation design for high-rise towers requires an
understanding of the building loads, geotechnical
conditions and regional seismicity. Structural details
such as basement depth, perimeter wall alignment,
column spacing, and column load variations are
required for a full understanding of the soilfoundation-structure system. Design aspects of
foundation type and depth of adjacent structures
often drive decisions for temporary excavation
support methods where deep basements are
desired. Temporary excavation support requirements
are considered in the foundation design, and may be
incorporated into the permanent foundation system.
Structural requirements create design and
construction challenges in the geology of the Mexico
City Federal District for design of high-rise towers.
Recent designs call for deeper basements and
construction directly adjacent to existing structures.
Since 1947, MRCE has been involved with Mexico
City projects, evaluating sites and designing
foundations. The firm has provided comprehensive
geotechnical and foundation design and testing
services since the 1990s in downtown and
surrounding suburban areas for major developers
(See Figure 1).
Example projects the authors have worked on
include geotechnical assessment of sites in Federal
District (e.g., Alameda Park site, Torre Mayor at

Chapultepec Parks east end, and Santa Fes


multiple developments,) and peer review services on
several other tall buildings in the Zona Rosa area.
The 55-story Torre Mayor (aka 505 Torre Reforma)
received the prestigious American Consulting
Engineers Council (ACEC) Platinum Award in 2004.

Figure 1. High-rise structures designed with MRCE


support.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

40

Foundation design and construction for high-rise Towers in Mexico City

2 GEOLOGIC SETTING AND SITE CONDITIONS

2.1 Tectonic and Seismologic Setting


Mexico City is subject to earthquakes and other
natural hazards such as floods and volcanic
eruptions. The region is located on a unique volcanic
high plateau at about 2240 m above sea level,
bounded by volcanic sierras, alluvial fans and plains
(Flores-Estrella et al., 2007).

earthquakes of magnitudes between 7 and 8, and


many frequent smaller events.
The ocean floor of the Cocos plate is subducting
beneath the continental edge of the North American
plate at a rate of about 6 cm/year. Figure 3 is a 3-D
image of the subducting slab developed by seismic
tomography (Caltech, 2015). The slab begins forcing
its way down beneath the continental crust at a
shallow angle, then levels off to nearly horizontal.
Below Mexico City it plunges steeply into the mantle
and ends abruptly at a depth of about 500 km.

Figure 3. Subduction image (modified from Caltech, 2015).

2.2 Geologic Setting the Ancient Lake

Figure 2. Seismotectonic setting and major tectonic plates.

The deep basement (Valley of Mexico) is faulted


and folded, contributing to basin seismic effects. The
plateau is located within the Trans-Mexican Volcanic
Belt (TMVB), a complex Tertiary and Quaternary
feature which crosses the country from the Pacific to
the Atlantic Oceans. Situated on a subduction zone,
the citys complex seismotectonic environment
consists of four major tectonic plates shown on Fig.
2 (North American, Cocos, Caribbean, and Pacific)
and a microplate (Rivera) that have generated

The site was originally a lake with islands connected


by causeways to surrounding higher ground. Spanish
colonists filled the lake between the islands in the
17th and 18th centuries.
The Federal District Paseo de La Reforma area
where high-rise structures under discussion are
located, lies within the Lower Transition Zone and
at the east edge of Largo Centro I (central lake zone)
on the area map of Fig. 4. This is the so-called Valley
of Mexico formed by volcanic materials interspersed
with alluvial deposits covered in the center of the
valley by lacustrine clays. Ground water pumped for
domestic supply has resulted in ground settlement,
causing widespread damage of structures and
infrastructure.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

DEMING P. et al.

Figure 4. Mexico City geotechnical zoning (after TGC).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

41

42

Foundation design and construction for high-rise Towers in Mexico City

2.3 Subsurface of Reforma (Zona Rosa)


As in most areas formerly occupied by the lakes, the
primary soil strata are soft clay deposits combined
with layers of stiff clayey silt and silty sands. A typical
subsurface cross section in the Zona Rosa area is
shown in Fig. 5 (Ovando-Shelley, et al., 2007).
Geotechnical performance is dominated by the
compressible Upper Clay Series clay and silt
deposits. The Upper Clay deposit (FAS) underlies the
man-made fill and is a crust of desiccated low
plasticity silty clay that contains volcanic ash, with
water content as high as 300% and Liquid Limit up to
400%.

Figure 5. Typical cross section in central Mexico City


(modified from Ovando-Shelley, et al., 2007). Designers
should develop a soil profile specific to their site.

The Upper Clay series is interlayered with compact


silty sands. These compressible deposits are
underlain by the Capa Dura a compact alluvial sand
layer at a depth of about 25-30 m, and the lightly preconsolidated lacustrine Lower Clay Series. The
Federal District along Paseo De La Reforma are
located lies within the Lower Transition Zone and
the east edge of Lago Cetro I. Performance of soils
in Lago Centro is dominated by the compressible
Upper Clay Series. At depths on the order of 33 m
the soil profile transitions to Depositos Profundos,
thick layers of medium compact silty sand. The depth
to bedrock within Lago Centro is so great that
structural support on bedrock is impractical.
2.4 Water Conditions
The shallow phreatic water table is on the order of 4
m deep; this is likely water perched in the Upper
Clay. Water extraction from deep sources has
lowered piezometric water levels in the deep
deposits. Prudent designs should assume the
shallow water table will be restored in the life of the
structure, and consider such hydrostatic forces will
act on the basement walls and base slab.

FOUNDATION DESIGN

3.1 Geotechnical Considerations


The Upper and Lower Clay series are not suitable for
high-rise building support and introduce substantial
risk in temporary excavation support, especially
adjacent to existing structures. Shallow mats in the
clay horizons also incorporate drilled shafts or piles
bearing on the Depositos Profundos silty sand
below 33 m.
There are two basic foundation systems, either
structural mat or mat on piles, or deep elements
alone as foundations independent of the base slab.
In either case, the base slab should be designed
as a pressure slab transferring hydrostatic uplift to
the columns, or with a permanent drainage system to
relieve hydrostatic pressure. The pressure slab is
preferred, as drainage systems require maintenance,
can increase effective stresses, and result in base
settlement and downdrag forces at the perimeter.
Design of the base mat as a structural mat for load
support takes advantage of the weight credit
provided by excavation and reduces dependence on
the load-deformation performance of individual
foundation elements.
For example, the 55-story tower at 505 Reforma
has four basement levels at 18 m deep, with a 2.5-m
thick piled mat bearing on the Capa Dura sands,
and drilled shafts bearing in the Depositos
Profundos below 33 m depth. The drilled shafts
were arranged at column locations and distributed
around the perimeter. Drilled shafts take advantage
of side friction in both clay and sand layers, and load
sharing between the mat and piers provides bearing
and overturning safety, and reduces settlement.
Newer structures that the authors are familiar with
involve greater basement depths for parking, some
requiring the removal of all of the Upper Clay Series.
The high-rise structures are supported on deep
foundations bearing in silty sand below 33 m depth,
without base mat contribution.
For deep basement designs, use of slurry
construction methods such as perimeter diaphragm
walls (slurry walls) and cast-in-place concrete drilled
shafts or Load Bearing Elements (LBE) at column
locations is desirable. These deep foundation
elements are constructed from the ground surface to
provide hydraulic head advantage to the slurry
methods, and because soil subgrade at lower levels
cannot support the construction equipment.
Use of diaphragm walls at the perimeter and topdown excavation support methods are popular, as
they alleviate the cost of temporary excavation
support bracing, provide rigid support of neighboring
properties and can be designed to prevent bottom
heave in soft clays. Slurry-supported construction
advanced from street grade increases contact
bearing stresses at the sidewalls of the deep

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

DEMING P. et al.

elements under fluid pressure of fresh concrete for


maximum friction benefit.
There remain many subsurface unknowns that
must be explored, considered, and controlled by the
design engineers to be confident that either the low
structure or high-rise tower is safe for occupancy for
50 years or more. These are explained below.
3.2 Structural Loads
Structural gravity and wind loads above grade can be
far easier calculated than the vertical and horizontal
reactions on the foundation elements in the soft clay
soils at basement level. Knowledge of soil
stratification, geotechnical strength and consolidation
characteristics, and piezometric pressure at depth
are required for reliable estimates of foundation
support capacity, downdrag forces, and structure
total/differential settlement. Load contributions are
not computed in one phase, but rather evaluated
several times during the design development, as the
interplay between final project structure conditions
and foundation performance are considered by the
designers.
Of particular importance is the influence of deep
excavation on adjacent structures. Many low-rise
structures supported on shallow footings or mat
foundations suffered severe damage and even
collapsed in past earthquakes. Following the violent
1985 Mexico earthquake of moment magnitude Mw
8.0 and a Mercalli Intensity of IX, local engineers
have an understanding of the importance of seismic
soil-structure interaction and are seeking more
advanced design guidance from international
specialized consultants. Using a more thorough
understanding of the soil-foundation-structure
system, designers can introduce more redundancy
by distributing the mass and stiffness to create
multiple load paths that are favorable to the safety
and performance of the structure.
3.3 Friction and Bearing for Deep Foundations
Even though elements constructed under slurry are
typically designed for side shear support without end
bearing, removal of sediment from the bottom is
mandatory. Low-density sediment is readily displaced
by concrete, and can coat the lower sidewalls of the
element to greatly reduce side friction. Soft sediment
trapped at the bottom reduces safety factor derived
from end bearing, introducing an undesirable
increase of risk to foundation performance. Where
building columns are supported on individual LBE
systems at wide spacing, the deformation of each
LBE is critical for adequate frame support and bottom
preparation should be carefully inspected.
An LBE constructed from grade can be designed
to incorporate an upper structural steel column
terminating in the concrete LBE below final subgrade,
or as LBE columns extending through the basement

43

space. In top-down construction, attaching floors and


a base mat to a concrete LBE is feasible with special
detailing. Shear dowels are drilled into or placed
through the LBE elements to support floor girders
and connect to a base mat (See Figure 6). The LBE
is often used in conjunction with perimeter diaphragm
walls since both are constructed with the same
equipment, reducing construction cost.
Drilled pier elements are a viable option, but
because of their smaller perimeter area, several
drilled piers must be clustered below a column that
require a structural cap or mat foundation to
distribute column loads to the deep elements.

Figure 6. Foundation mat top and bottom steel with LBE


connections before concrete pour.

3.4 Deep Foundation Recommendations


For high-rise tower foundation design, we provide
these general recommendations:
Deep foundations (piles, slurry walls, LBEs) with
and without soil grouting must be competently
designed to account for friction, downdrag, end
bearing and lateral resistance in the various soil
deposits that they are driven or drilled through.
The depths and consistency of soil strata
(especially the lake deposits) vary considerably,
and may be influenced by adjacent structures due
to group action and seismic lateral movements.
The geotechnical exploration should define both
shallow conditions which support adjacent
structures and deep deposits which are planned
for high-rise support.
Supporting foundation design with load testing of
trial deep foundation elements at the design depth
and constructed in the manner proposed for
production work is highly recommended. Load
testing with instrumentation to demonstrate actual
side shear and end bearing values which are then
applied to the final foundation geometry, provides

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

44

Foundation design and construction for high-rise Towers in Mexico City

confidence in foundation performance. If an


Osterberg load cell is used (See Figure 7),
cleaning the bottom of the test shaft to provide
competent bearing is required since the cell
obtains resistance from end bearing for stressing
the shaft side shear.
Use of post-grouting to improve deep foundation
element performance is feasible, but the
improvements should be demonstrated by testing.
Tip grouting is recommended as a proactive
remedy for potential soft sediment at the tip due to
slurry construction. Side grouting is desirable, but
may be ineffective in clay soil.

for moment and shear connections for belowgrade frame action.


4 FOUNDATION CONSTRUCTION

4.1 Support of Excavation Considerations


Deep excavations may be made with open-cuts and
sloped sides in soil, or they may be made with the
sides supported by a sheeting system (timber, steel
sheet piling or slurry walls) and bracing of various
types in a staged construction of a large site.

Figure 8. Couplers at block-out in LBE in top-down


construction.

Figure 7. Osterberg cells used at 505 Torre Reforma.

Construction methods must be carefully inspected


and slurry properties tightly controlled to prevent
reduction of sidewall resistance and prevent
sedimentation below support elements.
Diligent inspection and quality control testing are
needed during the execution of foundation work to
reduce owner and contractors risk and liability.
Floor keys should be blocked out for exposure as
construction progresses downward.
Couplers are needed when multiple segments of
preassembled cages or closely spaced rebar
cages are used in deep piles and slurry wall or
LBEs. Couplers (See Figure 8) are recommended

Open-cuts must have stable side slopes and be


above the ground water. Site conditions can be
determined by geotechnical borings and undisturbed
soil sampling to determine design properties. Where
the sides of the excavation are supported the support
wall must extend deep enough below the excavation
subgrade to prevent a base heave failure. A stability
analysis of the perimeter system should be
performed to maintain adequate safety factor.
With deep basement construction, slurry walls can
be extended deep below the temporary support
requirements to gain vertical load capacity and
enhance horizontal resistance. The walls also control
temporary seepage and base heave during
construction.
4.2 Construction of Deep Foundations
Top-down construction is a somewhat complex
technique which works well for high-rise towers with
deep basements. The use of top-down construction
of deep foundation basements offers many
advantages. The upper building may advance at the
same time as excavation below grade (Up-Down), so
long as the tower loads are supported by the
foundation method, and final foundations in staged
excavations are completed in concert with the tower

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

DEMING P. et al.

loading (e.g., Torre Mayor). The concept, illustrated


in Figure 9, allows schedule savings.
Slurry walls, internally supported by permanent
basement slabs, allow excavation spoils to be
removed through temporary openings in the floor
slabs, which are cast as the excavation progresses.
Top-down
construction
is
preferred
for
excavations deeper than 30 m with the following
advantages:
Elimination of temporary bracing and shoring
below grade if the permanent basement is cast as
excavation proceeds downward (See Figure 10).
High-rise structural work can commence early.
Imperfections in deep elements exposed by
excavation can be addressed before loads
increase above allowable.

Figure 9. Top-down basement construction illustrated and


example.

Figure 10. Staged temporary bracing of excavation at


Torre Mayor.

45

5 SEISMIC DESIGN

5.1 Local Code and Performance Based Design


Seismic design codes in Mexico have been
progressively developing for more than 7 decades
(Ordaz & Meli, 2004). Some important developments
adopted internationally include importance factors,
linear distribution of seismic forces with height,
dynamical analysis, and higher accelerations for soft
soil conditions.
The Mexico City Building Code published in 2004
is used as a model code in different municipalities
and states in the country (Alcocer & Castano, 2008).
Some key aspects include: (i) site effects in design
spectra are addressed using the predominant ground
period (See Figure 11) and (ii) two limit states
(service and collapse prevention) along with interstory drifts that reflect expected structural
performance. For essential structures, peer review
procedures are considered as risk reduction
measures.
The surge of high-rise construction at the turn of
the last century world-wide created a need for
performance-based design (PBD) approaches to
guide design for heights outside the range of building
code prescriptive provisions. The Pacific Earthquake
Engineering Research Center (PEER) responded to
this need by leading the Tall Buildings Initiative (TBI)
to develop engineering design criteria that will ensure
safe and usable tall buildings following future
earthquakes. As a result, the Guidelines for
Performance-Based Seismic Design of Tall
Buildings, was developed (PEER, 2010).
The Guidelines are an alternative to prescriptive
procedures such as those in ASCE7 and IBC
(International Building Code), intended for engineers
and building officials involved in seismic design of
individual tall buildings. The Guidelines consider
seismic response characteristics of tall buildings,
including relatively long fundamental vibration period,
significant mass participation and lateral response in
higher modes of vibration, soil-foundation-structure
interaction, etc. They recommend a sliding scale of
seismic hazard (or probability of an earthquake
happening during the design life of the building) with
corresponding performance objectives.
Globally, the PEER (2010) PBD Guidelines have
been extensively used for tall and mega-projects, and
in recent Mexico tall buildings in combination with the
local code. The PBD application in Mexico includes
three seismic hazard levels and corresponding level
of performance: Service Level (for a frequent event
with return period Tr = 43 years), Design Level (an
event that has reasonable probability to occur within
50 years of design life and return period of Tr = 125
years), and Maximum Considered Earthquake MCE
(extreme very rare event with Tr = 2475 years).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

46

Foundation design and construction for high-rise Towers in Mexico City

motion and large settlement on the order of 0.6 m


(end-bearing elements).
The 1985 earthquake led to tighter building codes
and the establishment of a Seismic Alarm System
which provides a 50-sec warning for any earthquake
greater than magnitude 6.0 occurring off the coast of
Guerrero or Michoacn.
5.3 Seismic Hazard

Figure 11. Predominant ground period (sec) in 2004


Mexico City Code (after Ordaz & Meli, 2004).

5.2 Historic Seismicity


Mexico has a long history of destructive earthquakes
and volcanic eruptions. Figure 12 shows events of
magnitude greater than 7.0 between 1900 and 2000.

The seismic sources affecting Mexico City are shown


on Fig. 13. The wave attenuation varies for each type
of source as shown on Fig. 14 for an assumed
earthquake magnitude of 7.5. This figure shows
attenuation of the spectral acceleration (SA) for a
structure with period T of 1 sec and distance from the
epicenter ranging from 0 to 500 km.
Uniform hazard predictions for each PBD design
level at bedrock can be made using the Probabilistic
Seismic Hazard Analysis (PSHA) approach originally
developed by Cornell in 1968. PSHA (See Figure 15)
combines all seismic sources, incorporating their
annual activity rate and the wave attenuation from the
epicenter to the site using a logic tree approach
where each assumption is assigned a weight to
produce uniform spectra at the bedrock or some stiff
interface layer.

Figure 13. Examples of shallow crustal (C) and intraslab (I)


seismic sources of Mexico.
Figure 12. Historic earthquakes with M>7 between 1900
and 2000 (ref: USGS).
1

0.1

SA (1.0 sec) : g

SA (1.0 sec) : g

On September 19 1985, an earthquake measuring


8.1 on the Richter scale (moment magnitude Mw of
8.0) centered between the states of Michoacn and
Guerrero, in the subduction zone off Acapulco, killed
more than 4000 people in Mexico City, 300 km away.
Double resonance coupling between the
earthquake shaking, soil layers, and buildings caused
intensity IX shaking, lasting up to 3 min in some
areas. Effects of soil amplification and topography
were pronounced in the lake area. Surface
foundation failures were observed and deep
foundations experienced reduction in shear strength
(friction elements) due to many cycles of strong

0.1

Atkinson & Boore (2003) - Worldwide Intraslab


Zhao
et al. (2006)
USGS Variant
Intraslab
Sources
Youngs (1997) USGS Variant - Intraslab Rock
Zhao et al. (2006) USGS Variant
Interface
Sources
& Boore
(2003) -Worldwide Subduction
Atkinson & Boore (2003) -Worldwide Subduction Atkinson
Abrahamson & Silva (2008)
Abrahamson & Silva (2008)
Boore & Atkinson (2008)
Boore & Atkinson (2008)
Shallow Crustal Sources
Campbell & Bozorgnia (2008)
Campbell & Bozorgnia (2008)
Chiou & Youngs (2008)
Chiou & Youngs (2008)

Atkinson & Boore (2003) - Worldwide Intraslab


Zhao et al. (2006) USGS Variant
Youngs (1997) USGS Variant - Intraslab Rock
0.01
Zhao et al. (2006) USGS Variant

0.01

0.001

7.5
MagnitudeM= =7.5

M = 7.5
1

Intraslab Sources
Interface Sources

Shallow Crustal Sources

0.001

10

SITE-TO-SOURCE
R : km

100

10

D I S T A NSCI TE E - T O - S O U R C E
R : km

500

100

500

DISTANCE

Figure 14. Wave attenuation of spectral acceleration for


magnitude 7.5 and structural period of 1 second.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

DEMING P. et al.

5.4 Site Response and Ground Motions


Site effects are significant in the deep soil profiles of
the Reforma area. One-dimensional equivalent linear
analysis can be used to capture soil effects and
supplemented as needed with nonlinear and twodimensional (2-D) models.
Figure 16 illustrates computed soil amplification of
the motion at the foundation elevation compared to
the motion at the bedrock. As in this case, it is typical
that the dominant natural soil periods are long, often
higher than 1 sec, and the response of the soil may
not be damped out depending on the thickness and
properties of the fine-grained soils. Additional ground
motion effects due to topography depend on the site
location with respect to the valley.

Figure 15. Steps of PSHA (from Cornell 1968).


3.5
Ground Surface - 2475 yr

AF = 3.0

Amplification Factor, AF

3.0

Average AF

2.5
2.0
1.5
1.0
0.5
0.0
0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

47

There are several approaches for incorporating


Soil-Foundation-Tower interaction in the PBD
analysis. Listed in order of simplicity:
Fixed-base: The superstructure is assumed to be
fixed at the surface and the input motion is the
free-field (FF) motion at the ground surface (Fig.
17a). The results do not consider interaction and
are a baseline for more detailed analyses.
Foundation model (service earthquake): The input
motion at the bottom of a rigid bathtub (Fig 17b)
can either be the FF or the FIM (Foundation Input
Motion), modified to account for kinematic
interaction. This approach does not account for s
soil springs and the subterranean levels. Mass
should include the mass of the tower below grade.
Foundation model (MCE): Springs and dashpots
representing foundation-soil interaction along the
foundation sides and base are included (Fig. 17c).
Input motions are applied via a rigid foundation
frame as either FF or FIM. Nonlinear response
history analysis is performed using 3D model with
the ground motion applied at the foundation base
or directly through the distributed soil springs.
The importance of incorporating the foundation
compliance with equivalent springs depend on the
foundation system selected. In Mexico City, there
have been tall buildings whose foundations are mats
structurally designed to transfer the load to the upper
soil layers, often complemented by deeper
foundations to ensure safety and account for
buoyancy effects.
In this case, the equivalent springs and their
distribution below the mat can result in a longer
structural period as compared to the fixed-base
model (Fig. 17a). In smooth, code-type spectra, this
translates to a lower base shear but to also larger
computed deformations of the structure, since it is
more flexible. When a slurry wall acts also as the
perimeter foundation wall, the interaction between
the wall and the surrounding soil should be properly
represented by equivalent springs, especially since
these springs may vary in stiffness for each soil layer.
The complete soil-foundation system should be
studied with the input applied at the bottom of the
rigid mat (Fig. 17c).

Period, T : s

Figure 16. Amplification of seismic motion at ground


surface from the soil layers above bedrock.

It is essential that the site is characterized sufficiently


with dynamic measurements of soil shear wave
velocity and laboratory testing.
5.5 Foundation-Tower Interaction
The physical problem of Soil-Foundation-Tower
Interaction is illustrated on Fig. 17 (top) and various
model simulations are illustrated on Fig. 17 (bottom).
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

48

Foundation design and construction for high-rise Towers in Mexico City

Ref: Nikolaou (2008)

Figure 17. Soil-Foundation-Tower Interaction: (top) actual


problem (Nikolaou, 2008); (bottom) model simulation
options (Klemencic et al., 2012).

In other designs, the load transfer relies mostly on


LBEs that transfer loads to deep competent soil
layers, while the slab acts essentially as slab on
grade without providing stiffness or load transfer. In
this case, the stiffness of the LBE controls the
response and shift of the structural period. This may
not be significant for very stiff rigid elements and the
simplified model of Fig. 17b can be applied with
some modifications on the input ground motion that
should be filtered for kinematic effects (Klemencic et
al., 2012). The complete soil-foundation system
should be studied with the input applied at the tip and
along the shaft of the deep LBEs (Fig. 17c).
The responsibility of the analysis and design of the
foundation systems is strongly dependent on good
collaboration between the structural and geotechnical
engineers who often go through several iterations to
ensure compatibility of deformations and strains
between the superstructure and substructure models.

design. Testing must also consider the construction


methods, especially where slurry stabilization is used.
The high skin friction provided by the upper clays
challenges static load testing of the deep sands, but
enables the Osterberg load cell method to develop
sufficient reaction for measurement of the unit friction
in the deep sand. The Osterberg load cell has been
in very few developments in Mexico City. Future
Needs: More available test data defining the unit
friction capacity of the deep sands being engaged in
friction for high-rise structure support.
Regional seismic activity is high, and the great
depth to bedrock with soft soil and clay deposits can
result in an elongated period of the soil-foundation
system. Soil-structure interaction analysis may not
lead to a reduction in inertial forces for tall structures,
as the soil may also vibrate in long periods. The PBD
design approach has become a standard practice for
high-rise designs in Mexico City, in combination with
the local Codes. Future Needs: Better agreement on
assumptions made for hazard evaluations, such as
selection of seismic sources and attenuation models.
Foundation-tower interaction analysis should be
performed in progressive stages as design details
are developed. For deep excavations which remove
the upper clays, the complete soil-foundation system
should be studied with the earthquake excitation
applied at the bottom of the rigid mat, or at the tip
and sides of the deep foundation elements where a
mat is not constructed.
High-rise construction loads result in high bearing
stresses on mats or high shear stress at deep
element interfaces. These loads are also applied to
the soil adjacent to the perimeter diaphragm wall,
which typically performs as a foundation in bearing.
Future Needs: Static and dynamic interaction of
bearing support soil should be considered where new
structures are in close proximity to existing, and new
foundations are constructed several stories deeper
than adjacent foundations.
REFERENCES

6 CONCLUSIONS AND FUTURE NEEDS


The construction demands introduced by deep
excavations into the soft upper clay deposits of
Mexico City, and the need to support excavations
adjacent to structures supported on shallow
foundations, allow permanent diaphragm walls and
top-down construction methods to be competitive
with temporary excavation support methods.
Mat foundations with shared deep element support
at columns, and deep element support with only
pressure slab at grade are popular foundation
options due to the challenging subsurface conditions
of Mexico City.
The use of pile load testing to measure unit skin
friction is needed for safe and practical foundation

Alcocer, S.M. & Castano, V.M. (2008). Evolution of


codes for structural design in Mexico, UNAM
Avilsa, J. & Prez-Rochab L.E. (2010). Regional
subsidence of Mexico City and its effects on
seismic response, Soil Dynamics & Earthquake
Engineering, 30(10):981-989
Caltech (2015). The unusual case of the Mexican
subduction zone, CalTech Tectonics Obervatory.
Chvez M. y Alcntara L. (1990). Interaccin sueloestructura en estaciones acelerogrficas de la
ciudad de Mxico, Memorias XV Reunin
Nacional de Mecnica de Suelos, SLP, 1: 69-76
Cornell, C.A. (1968): Engineering Seismic Risk
Analysis, BSSA, 58:1583-1606

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

DEMING P. et al.

Flores-Estrella, H. et al. (2007). Seismic response of


Mexico City Basin: A review of 20 years of
research, Natural Hazards, 40:357-372
Hadley, P.K. et al. (1991). Subsoil geology and soil
amplification in Mexico valley. Soil Dynamics &
Earthquake Engineering, 10(2):101-109
Klemencic, R., McFarlane, I.S., Hawkins, N.M.,
Nikolaou, S. (2012). Seismic Design of Mat
Foundations, NIST/NEHRP Design Brief No. 7.
Nikolaou, S. (2008). Site-Specific Seismic Studies
for Optimal Structural Design, Structure, Feb.
Ordaz, M. & Meli, R. (2004). Seismic design and
codes in Mexico, 13th World Conf. on Earthquake
Engineering, Vancouver, Canada, Paper No. 4000
Ovando-Shelley, E., Ossa, A., Romo, M.P. (2007).
The sinking of Mexico City: Its effects on soil
properties and seismic response, Soil Dynamics
& Earthquake Engineering, 27:333-343
PEER (2010). Guidelines for Performance-Based
Seismic Design of Tall Buildings, Tall Buildings
Initiative (TBI)
Stone, W.C. et al. (1987). Engineering Aspects of
9/19/1985 Mexico Earthquake, NIST, NBS BSS16
Tamaro, G.J. et al. (2000). "Design & construction
constraints imposed by unique geology in New
York," DFI 8th Int. Conf., NYC
TGC Geotecnia, MRCE (1994). Geotechnical and
environmental investigation for the Chapultepec
Tower project, Mexico, DF [now Torre Mayor].

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

49

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Deep foundations in Mexico City soft soils


Cimentaciones profundas en suelos blandos de la Ciudad de Mxico
Gabriel AUVINET-GUICHARD1 & Juan-Flix RODRGUEZ-REBOLLEDO 2
1

Instituto de Ingeniera, UNAM, Mxico


2 Universidade de Brasilia, Brasil

ABSTRACT: A general overview of the different solutions to the problem of foundation of buildings on highly
compressible soft soils such as those found in Mexico City valley is presented. Special attention is given to deep
foundations systems that were developed to control settlement or protruding in consolidating soils. Recent techniques for
modeling the behavior of deep foundations are also reviewed.

1 INTRODUCTION
The lacustrine soils of Mexico City are famous
worldwide for their high water content, poor shear
strength and high compressibility [35]. Since the
founding of the city, Mexican engineers became
aware that, in many cases, the techniques used in
other countries were not directly applicable to the
exceptionally
difficult
geotechnical
conditions
prevailing in the area. They had to innovate. This was
not an easy task and some foundation designs
proved unsatisfactory due to wrong reasoning,
forgotten factors or overly optimistic assumptions.
The most severe judges of the quality of the
foundation systems were the earthquakes that affect
periodically this area; the harshest lessons, but also
the most useful, were derived from seismic events
and especially those of 1985 [5].
In this paper, an overview of the contributions of
many researchers, consultants and builders to
foundation engineering in the soft soils of Mexico City
is presented, emphasizing deep foundation systems
and special solutions that have been proposed to
overcome the specific difficulties of the area. Recent
contributions of numerical modelling and full-scale
observations to a better understanding of foundations
behavior and improved design criteria, most of which
have been included in Mexico City building code, are
exposed.
2 GEOTECHNICAL CONDITIONS
Until the end of the XVIIIth century, the valley of
Mexico was a closed basin with a number of shallow
lakes, amongst them: Texcoco, Xaltocan and Chalco
lakes. The capital of the Aztec empire, Tenochtitlan
was built on a small island of the Texcoco Lake. The

valley became an open basin when the Nochistongo


cut, a channel 7km long and up to 50 deep, was dug
by hand between 1637 and 1789. Progressively, the
lakes were drained, initially through the Nochistongo
cut and later through the Tequisquiac tunnels (1900
and 1952) and the Deep Drainage System (Emisor
Central, 1975), and practically disappeared. A large
part of the city was then built on lacustrine sediments
which are highly compressible volcanic soft clays
interbedded with layers of silt and sand and sandy
gravels of alluvial origin.
2.1 Soil profile
As shown in Fig. 1, the urban area of Mexico City can
be divided in three main geotechnical zones: Foothills
(Zone I), Transition (Zone II) and Lake (Zone III), as
defined in the present building code [23]. In Zone I,
very compact but heterogeneous volcanic soils and
lava are found. These materials contrast with the
highly compressible soft soils of Zone III. Generally,
in between, a Transition Zone (Zone II) is found
where clayey layers of lacustrine origin alternate with
erratically distributed sandy alluvial deposits. The
main difficulties for foundation of high-rise buildings
are encountered in Zones II and III. An updated
zoning will be available in 2015.
In Fig. 2, a typical soil profile corresponding to the
Lake Zone is presented. The water table is close to
the surface. Three clayey layers are to be
distinguished, denominated upper clay formation
(Formacin Arcillosa Superior, FAS), lower clay
formation (Formacin Arcillosa Inferior, FAI) and
deep deposits (Depsitos Profundos, DP). The clays
of FAS are separated from FAI by a hard layer (Capa
Dura, CD), a sandy clayey stratum, some 3m thick,
found at a typical depth of 30 to 35m. Generally, FAS

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

52

Deep foundations in Mexico City soft soils

is covered by a desiccated crust and/or an artificial fill


with a thickness varying from a few decimeters to
several meters. Average values of index properties

are presented in Table 1 for a typical borehole in the


Lake Zone.

Figure 1. Geotechnical zoning of Mexico City [23].

Table 1. Typical average values of index properties in


Lake Zone (Borehole Pc-28, Marsal, 1975).
PROPERTY

FAS

CD

FAI

Water content, %

270

58

191
288

Liquid limit wL, %

300

59

Plastic limit, wP, %

86

45

68

Density of solids, Ss

2.30

2.58

2.31

Initial void ratio, e0

6.17

1.36

4.53

Unconfined compressive
strength, qu, kN/m2

85

24

160

Spatial variations of soil properties in the lacustrine


zone have been registered in a data base consisting
of more than 10,000 pits and boreholes. This data
base was incorporated into a Geographical
Information System focused on geological and
geotechnical features.
Distribution of soil properties within any specific
area of interest can be assessed from the database.
Virtual soil profiles and cross sections can be defined
using geostatistical estimation or simulation
techniques. As an example, Fig. 3 shows the
variation of soil water content within FAS along an EW axis.
Figure 2. Soil profile in the Lake Zone of Mexico City
(Marsal, 1975).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

53

AUVINET G. et al.

Figure 3. Spatial variations of water content.

a)

2.2 Regional subsidence

2.3 Seismic site effects

150

200

250

300

350

Pore Pressure, kPa


50

100

150

200

250

300

Crust
Crust
5

5
Hydrostatic
Oct-80

Hydrostatic

Jun-92

10

Dec-94

10

Aug-05

Feb-97

Upper clay
formation
20

Depth, m

Oct-99

15

15

Upper clay
formation
20

25

25

30

30

35

Pronounced site effects leading to amplification of


earthquakes affecting the basin of Mexico must also
be taken into account in foundation design in the lake
zone [52].

100

Depth, m

Exploitation of underground aquifers for supplying


potable water to the growing population, causes a
progressive depletion of the piezometric profile in
most of the valley (Fig. 4). As a consequence,
effective stresses within the soil increase and induce
consolidation. Mexico City is thus affected by a
regional subsidence that, in some locations has
reached an accumulated value of 13.5m since 1862
[14]. Recent data show that the rate of subsidence
tends to decrease in certain areas. However, in newly
developed urban zones, such as the eastern parts of
former lakes of Texcoco, Xochimilco and Chalco, the
consolidation process is only in its first stage and the
rate of subsidence can be as high as 40cm per year
[14]. In transition zones between soft and firm soils,
the subsidence induces differential settlements and
soil fracturing, causing severe damages to
pavements and small constructions [15,16].

b)

Pore Pressure, kPa


50

Hard Layer

Hard Layer
Lower clay formation

35

Figure 4. Typical piezometric profiles in the lake zone of


Mexico Valley.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

35

54

Deep foundations in Mexico City soft soils

3 TYPES OF FOUNDATIONS
3.1 Design criteria
Foundations of buildings in the lake area of the basin
of Mexico must be designed to contend with the
severe conditions described in the preceding
paragraphs.
Using the terminology of the Building Regulations
for the Federal District [23], foundations must provide
adequate security with respect to multiple limit states:
a) Failure limit states: floating, local or general
plastic flow of soil under the foundation and
structural failure of piles or other foundation
elements.
b) Service limit states: average vertical
movement, settlement or emergence with
respect to the level of the surrounding
terrain, average inclination and differential
deformation.

Safety with respect to these limit states must be


guaranteed for short and long term static loads but
also for accidental actions, particularly in seismic
conditions.
It should be recognized that, with few exceptions,
before the 1985 Mexico City earthquake, designs
almost exclusively focused on control of settlement
or emersion of foundations under static conditions.
The concern for controlling these movements led
designers to use many types of foundation including
traditional ones but also many special systems, some
of them very ingenious.
3.2 Conventional foundations
The solutions traditionally adopted for foundations of
buildings in Mexico City range from isolated or
continuous footings or slabs for buildings with few
levels, to concrete point bearing piles for high-rise
buildings (Fig. 5).

Figure 5. Conventional foundations on soft soils in Mexico City.

3.2.1 Surficial and compensated foundations


Foundation on masonry footings or general raft,
sometimes with short wood piles, was the first
system tested by the founders of the city, with very
little success at that as attested by the spectacular
problems registered in the foundation of Templo
Mayor, the main pyramid of the Aztecs, and of many
heavy colonial temples such as the Metropolitan
Cathedral. It is now accepted that surficial
foundations on footings or surficial mats are only
suitable for very light constructions occupying a
relatively small area. It must be taken into account
that a load of only 20kPa applied on a large area of
the lake zone can be expected to induce a total
settlement close to 1m with differential settlements of
about 50cm. Moreover, these foundations are
vulnerable to movements induced by adjacent
buildings.
Some of the problems faced when using surficial
foundations can be managed recurring to
compensated or floating foundation. The wellknown compensation technique consists of designing

the foundation, generally a box-type structure, in


such a way that the mass of excavated soil will be
comparable to the mass of the building [20].
Theoretically, if both weights are equal, the soil below
the foundation is not submitted to any net stress
increment and no significant settlement should
develop. When the weight of the soil is smaller than
the weight of the building, the foundation is partially
compensated; in the opposite case, it is
overcompensated.
In practice, even perfectly compensated
foundations undergo some absolute and differential
vertical movements due to soil elastic deformation, to
soil disturbance during construction and to static soilstructure
interaction
thereafter.
Furthermore,
constructing this type of foundation is not
straightforward since a deep excavation in soft soil is
generally required with the associated problems of
stability of earth slopes or support systems and to
bottom expansion or failure. Water tightness of the
foundation is also a critical factor for compensated
foundations; in many cases, this type of foundation

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

AUVINET G. et al.

must be equipped with a permanent pumping system


to control infiltrations.
3.2.2 Point-bearing piles
Precast or cast-in-place end-bearing piles embedded
in a deep hard stratum are an apparently obvious
solution for foundations on soft soils. Moreover, this
technique has proven to be more reliable than other
types of foundation in seismic conditions in Mexico
City. However, foundations on point-bearing piles
may present some substantial problems and their
design may face severe difficulties.
The bearing capacity of the hard layer in which the
piles rest is a first source of uncertainty. The
shortcomings of classical analytical methods for
evaluating this capacity have long been recognized.
Most of them assume rigid-plastic behavior of the soil
ignoring the essential role of soil deformability.
Bearing capacity estimations thus tend to be based
principally on in situ tests (cone penetration test,
pressuremeter) or on loading tests. Heterogeneity of
these hard strata is difficult to assess and may
originate tilting of buildings with such a foundation.
In consolidating soils, negative skin friction
develops on the pile shaft [4]. Moreover, an apparent
emersion of the structure with respect to the
surrounding area is generally observed (Fig. 6) and
damage can be induced in adjacent buildings
supported by other types of foundation. Consolidation
has also the effect of separating the slab of the
substructure from the soil. In that condition, the head
of the piles is no longer confined and can be
structurally vulnerable to the combined effect of
seismic overturning moment and base shear [6].

55

1) shear stress developed on the shaft of a pile cannot be larger than the limit soil shear strength.
2) limit shear stress can only be reached when
the soil attains the corresponding required
shear deformation.
3) axial force developed in a pile due to skin
friction within a pile group cannot be larger
than the weight of the soil located within the
tributary area of the pile.
4) unloading stresses induced by negative skin
friction within the soil cannot be larger than
those that are sufficient to stop the consolidation process that originates the skin friction
in the first place.
Curiously, many of the methods available to take
into account the negative skin friction do not consider
all of the above conditions, especially the last one.
These factors can easily be taken into account using
numerical (finite element) modelling [10, 50].
As mentioned above, foundations on point-bearing
piles presented generally an acceptable behavior
during the 1985 earthquakes. However, some cases
of structural damages in the upper part of the piles
were detected (Fig. 7). They were attributed to load
concentrations in the perimeter of the structure due
to overturning moment and base shear force.

Figure 7. Seismic damage in the upper part of a pointbearing pile (1985).

3.3 Special foundations


Figure 6. Apparent protruding of a foundation on endbearing piles in Mexico City.

3.3.1 Objective

As recognized in Mexico City building code, when


estimating the downdrag force induced on piles by
negative skin friction, the following considerations
should be taken into account:

Special deep foundation systems have been


developed with the principal objective of avoiding
both excessive settlement and apparent emersion
associated to consolidation of the upper clay
formation (Fig.8). Some systems also allow
controlling the load transmitted to each pile.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

56

Deep foundations in Mexico City soft soils

Figure 8. Some special foundation systems for soft soils in Mexico City.

3.3.2 Types of special foundations


The different systems all have in common the
inclusion in the piles of a fuse (an element

presenting large deformations when a critical load is


exceeded) allowing the construction to follow the
regional subsidence

Table 2. Principal types of special foundations.


Type

Fuse in lower part of pile

Friction piles

Piles with penetrating tip

P3 piles

Telescopic piles

Fuse in upper part of


pile

Negative skin friction piles

Control piles

Overlapping piles

In Table 2, the principal systems have been


regrouped according to the position of this element
(in the upper or lower part of the pile, or both). The
type of fuse used is characteristic of each system; in
some instances, the fuse is the soil itself.
Another solution, not included in the above table,
consists of using piles placed within a flexible casing
[55]. These piles are designed to avoid overloading
of point bearing piles by negative skin friction.
a) Friction piles
Friction piles are generally used to transfer stresses
induced by shallow or partially compensated
foundations to deeper, less compressible layers of
the subsoil, and to reduce settlements. Not so often,
they constitute the main foundation system and the
stability of the structure is dependent on the bearing
capacity of the piles. A clear distinction must be

established between these two types of design (Fig.


9; [5])
Type I: Design in terms of bearing capacity
In this first type of design, the number and
dimensions of the piles are selected with the aim of
guaranteeing that they will be able to support the load
from the structure under static as well as dynamic
conditions, with a safety factor generally larger than
1.5. In areas affected by regional subsidence, this
type of friction pile is submitted to complex loading
conditions (Fig 9). It has been shown [59, 46, 63, 4]
that negative skin friction can develop on the upper
part of the piles while positive friction develops in the
lower part. A "neutral" level can then be defined
between these two zones, where no relative
displacement occurs between soil and piles.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

AUVINET G. et al.

When the neutral level is in a low position (large


number of piles or high strength of the lower layers),
the downdrag force induced by negative skin friction
may lead to significant compression stresses in the
piles. Moreover, with time, the head of the piles can
be expected to protrude from the surrounding ground

57

due to consolidation of the soil located between the


surface and the neutral level, Fig. 10.

Figure 9. Friction piles.


Train

Road

Pavement
Slab

Slab apparent
protrusion

Friction
Hard layer
piles
Figure 10. Apparent protruding of a footing on friction piles (Type I design).

When this design philosophy is adopted, the


bearing capacity of piles must be estimated taking
into account the possibility of group behavior. When
the density of piles is high, soil friction available on

the perimeter of the pile group plus its base capacity


can in effect be smaller than the sum of the
capacities of individual piles [45].

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

58

Deep foundations in Mexico City soft soils

For piles working in the conditions indicated in Fig.


9a, settlements cannot be calculated by simplistic
methods such as the well-known "2/3 rule" equivalent
loading [42]. Depending of the position of the neutral
level, the foundation can in fact present either
settlement or emersion. Details of a more realistic
method to estimate foundations movements adapted
to these conditions were presented by Resndiz and
Auvinet [46] and recent developments of this model
are summarized in paragraph 4.1.
Type II: Design in terms of soil deformation
In this case, only a limited number piles is used with
the principal objective of reducing the settlements of
a partially compensated foundation (compensated
foundations with friction piles; [61, 62, 63, 64]). Since
the number of piles is low, the neutral level generally
coincides with the piles cap (Fig 8b). In that case,
positive friction is mobilized along the full length of
the piles, and the piles are in a permanent failure
state, which justifies the name of creep piles that
they were given by some authors [30].
Problems similar to those discussed for
compensated foundations may occur. Reliability is
low against excessive settlements in static conditions
[8].
Without any doubt, foundations of this type were
those that suffered most damages during the 1985
earthquake. 13% of all buildings between 5 and 15
stories, most of them on compensated foundation
with friction piles, experimented settlement, tilting,
punching of the soil (Fig. 11) and, in one case, total
failure.

during a seismic event due to widening by lateral


seismic forces of the perforation where the pile was
installed and be further reduced by remoulding of the
soil as cyclic shear stresses develop at the interface
between soil and pile. This was confirmed by
observations on an instrumented bridge on a box
type foundation with friction piles [39, 40].
Instrumentation included load cells in piles and
pressure cells below the slab of the substructure, as
well as piezometers and accelerometers. A transfer
of loads from the piles to the slab during earthquakes
was clearly observed.
Full scale experiments performed by Jaime et al.
[32] have shown that piles fail when the combination
of sustained plus cyclic loading exceeds the static
bearing capacity during more than ten cycles. When
the total loading exceeds this value by more than
20%, the subsequent sustained bearing capacity
decreases to a value as low as 50% of the static
capacity, while a penetration of the pile of 10 cm or
more is observed.
In the laboratory, some direct shear tests of the
soil-concrete interface have also been performed
[43]. The results show that static friction decreases
significantly after cyclic loading
For this type of design, it is thus commendable to
ignore the contribution of the piles to the global
bearing capacity. The bearing capacity to be
considered under seismic conditions should be
merely the capacity of the soil to resist the slab
contact pressure. The presence of the piles should
only be taken into account in static settlement
estimations [23].
There has been a number of proposals aiming at
increasing the efficiency of friction piles by modifying
the shape of their cross section (triangular, H, etc.).
Jaime et al. [32] have shown that this is generally not
achieved. Among the research work aiming at
improving friction piles, attempts to develop high
adherence electro-metallic piles using electroosmotic
treatment should also be mentioned [56, 57]
b) Pile with penetrating tip.

Figure 11. Failure of a foundation on friction piles (Type II


design) during the 1985 earthquake.

This type of pile [44] was conceived to increase


the bearing-capacity of friction piles with a controlled
contribution of the piles point. The diameter of the
point is smaller than the rest of the pile in order to
facilitate penetration in the hard layer under the
combined effect of loading and negative skin friction
and to avoid protruding. The point can be made of
reinforced concrete [44, 21] or steel [45]. In the latter
case, the bearing capacity of the pile can be better
controlled by using a point with a pre-established
failure load. Flexibility of the point has been found to
be a problem during installation of piles.

A creep pile cannot be expected to absorb cyclic


loading during earthquakes, since soil-pile adherence
is already fully mobilized, and can even decrease
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

AUVINET G. et al.

c) Negative skin friction piles


Those are simple point-bearing piles that
penetrate freely through the foundation slab [17].
They can contribute to reduce significantly the
settlements due to the negative skin friction that
develops on the shaft of the piles under the
combined effect of the structural load and the
consolidation of the clay layer. Finite element
modelling of this type of piles has been presented
[50]. Spacing of piles appears to be the most
significant design parameter for this type of
foundation.
d) Control piles
The so-called control piles are similar to the
previous ones but are equipped in their upper part
with a mechanism that controls the load received by

59

each pile (Fig. 12). Piles can also be unloaded by


removing the mechanism in order to correct any
tilting of the building. Close to a thousand buildings
are equipped with this system in Mexico City. Those
systems have sometimes been installed during the
life of the structure as part of an underpinning
process [28, 29, 7]. The different available control
mechanisms have been reviewed by different authors
[36, 19, 1, 49]. In Table 3 a list of the best known
systems is presented. In seismic conditions, some of
these special systems can be vulnerable and suffer
damage ranging from simple tilting to total collapse.
Lack of maintenance can also be a problem. Several
proposals have been made to improve the design of
control piles [1]. Overturning of the loading frame can
be avoided using a new type of anchors. The
mechanism can also be adapted to absorb tensions.

Figure 12. Control mechanism (Gonzlez Flores system).

Table 3. Principal types of control mechanisms for piles.


Mechanism

Reference

Loading frame with deformable wood


cubes

Gonzlez Flores, 1948; Salazar Resines, 1978

Loading frame with jack and


automatic relief valve

Pilatovsky, mentioned by J.J. Correa, 1980

Metallic tensors

Gonzlez, 1957, mentioned by Aguilar, 1990

Metallic cap

Aguilar, 1960, mentioned by Aguilar, 1990

Loading frame with flat hydraulic jacks

Streu, 1963, mentioned by Correa, 1980 and Aguilar,


1990

Sand confined within a capsule

Creixell and Correa, 1975, mentioned by Aguilar,


1990

Energy dissipater

Aguirre, 1981; Resndiz, 1976

Mechanical system of self-control

Jimnez, 1980

Mobile wedge

Girault, 1986, mentioned by Aguilar, 1991

Communicating hydraulic jacks

Zamora Milln, mentioned by A. Rico A., 1991

Constant friction cell

Tmez, 1988

Cell with teeth for transmission of


tensions

Rico, 1991

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

60

Deep foundations in Mexico City soft soils

e) Telescopic piles

g) Rigid inclusions

These are tubular piles with a piston-like cylindrical


point [18]. The tubular portion of the pile is partly filled
with sand. When the sand reaches a certain level, an
arching effect develops and both parts of the piles
become solidary. If necessary, sand can be removed
to free the point and avoid emersion of the
foundation.

During the last decade, foundations on rigid


inclusions have been used for a number of housing
complexes in Mexico City [12, 13, 65, 66, 67, 68].
They consist of a simple slab resting on a layer of
natural or artificial granular material (distribution
layer) which in turn rests on structural elements,
generally made of plain concrete or steel, previously
placed in the ground (Fig. 13). In Mexico City, the
distribution layer is not strictly necessary since a
competent dry crust is found in the first meters of the
soil profile. These "inclusions" constitute an
economic foundation system for control of
settlements. They have been used successfully for
mid-rise buildings, which are only marginally affected
by earthquakes [69]. To avoid disposal of large
volumes of extracted soil, inclusions can be built
using a displacement auger.

f) Overlapping piles
This type of foundation [24, 25, 26] uses conventional
friction piles (A Piles) combined with negative skin
friction piles (B piles) lying on the hard layer. This
arrangement reduces the stress increments in the
soil and the corresponding settlements. Apparent
emersion can also be avoided. This system has been
used for the foundation of buildings and oil tanks in
Mexico City [38]. It looks like this solution, used in
Mexico City for many years, has been rediscovered
recently by other authors [34].
Dried
crust

Mat

Extra perimeter
inclusion row
2m

19m

Upper clay formation

0.4m circular
inclusions

8m

Hard Layer

Figure 13. Typical foundation on rigid inclusions.

h) Foundations on cast-in-place walls and structural


cell
Foundations on barrettes and cast-in-place walls are
increasingly common in Mexico City. Special mention
should be made of foundations consisting of four
cast-in-place walls forming a structural cell as shown
on Fig. 14 [37]. In this type of foundation, advantage
is taken of the embedment of the cell into the soil to
resist seismic actions. The width of the cell can then
be significantly smaller than the one of a classical
box-type foundation on friction piles, an important
advantage in an urban context.

ADVANCES IN MODELLING OF DEEP


FOUNDATIONS IN CONSOLIDATING SOILS

Different types of models have been developed along


the years to represent the complex behavior of deep
foundations in soft consolidating soils.

4.1 Analytical model [46, 3, 50]


This model is useful to estimate the stresses in the
complex field generated by loads caused by positive
and negative skin friction on piles as described in Fig.
8. From the vertical stress distribution and taking into
account the results of odometer tests, local

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

AUVINET G. et al.

settlement can be estimated. Two versions of this


model are available.
The first one [46] is a simplified model where
equivalent horizontal areas with uniform loading at
different depths are used to represent the positive

61

and negative skin friction acting on the piles shaft


(Fig. 15).

Figure 14. Foundation on structural cell.

Figure 15. Simplified load distribution model for friction piles [46].

The second model [3] computes explicitly the


stresses induced by each individual pile as well as by
the contact pressure between the slab and the soil,

using elasticity theory (Mindlin and Geddes


solutions).
Due to its simplicity, the 1973 model is still one of
the most commonly used by geotechnical designers.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

62

Deep foundations in Mexico City soft soils

An important improvement of this model has been


introduced recently. To assess the negative skin
friction in the upper part of a pile, consideration 4) of
paragraph 3.3.2 is explicitly taken into account.
According to this consideration, negative skin friction
developed on the upper part of the pile cannot
exceed the seepage forces due to piezometric
depletion acting on the soil contained within the
tributary area around the pile.
The procedure implemented in software MICRA
[41] is thus as follows:
Neutral level elevation (z0), is defined solving the
following equation:

Q
FN FP C P
NP

FN C f

z0
Df

The local neutral level for these piles will tend to


occupy a higher position than for piles located within
the group. Larger settlements should then be
expected in the foundation perimeter, a fact
confirmed by observations in many buildings.
Introducing a downdrag force FN equal to z0 AT
in the model proposed by Resndiz and Auvinet (Fig.
16) allows evaluating the local settlement (settlement
at the elevation of the neutral level elevation) at
different times in the future corresponding to different
hypotheses regarding the evolution of pore pressure
depletion profile.

(1)

where

sum of permanent actions plus variable actions


with medium intensity;

NP number of piles;

FP C f

D f LP
z0

Positive friction equal to the limit

shear strength that can be developed from z0 to


depth of pile tip (Df + LP);

Cp
LP
Df
FN

estimated point bearing capacity


length of piles;
depth of substructure (slab)

negative skin friction estimated as follows:


C z0
f D f
FN mn.
(2)

z0 AT

where:

Cf

z0
Df

estimated limit shear strength that may deve-

lop on the pile shaft from depth Df to z0.

z0 effective stress increase at depth z0 (ignoring


the presence of piles) generated by the pore
pressure depletion estimated for the future at a
given date. Note that z0 AT is the resulting

Figure 16. Considerations for negative skin friction on a


pile.

Comparing this local settlement to the estimated


or computed regional consolidation of the soil
between the tip and the bottom of the piles, it is
possible to conclude whether the foundation will
settle or protrude with respect to the surrounding
terrain in the future (Fig. 17). An optimum design
should aim at minimizing both settlement and
protruding during the life of the structure.

force at the neutral level elevation due to vertical


seepage forces acting on the soil around the
upper part of the pile.

AT tributary area of each pile.


Perimeter piles require special attention since the
tributary area of these piles is not limited horizontally.
In that case, the skin friction can be considered as:
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

AUVINET G. et al.

63

4.2 Numerical models

Figure 17. Evolution of general subsidence,


settlement and resulting apparent movement.

local

Advances in numerical modeling of the static


behavior of pile groups and inclusions in soils
subjected to regional consolidation have been
presented recently [50, 51]. Parametric studies were
performed using 2D, axisymmetric and 3D numerical
models of isolated piles and pile groups (Fig. 18),with
advanced constitutive soil constitutive laws models
(Soft Soil or S-Clay1, [60]).
To simplify 3D models it has been found useful to
resort to slice models taking advantage of
symmetry and approximate plane strain conditions
prevailing within piles groups (Fig 19).

Figure 18. 3D model of a group of piles [50].

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

64

Deep foundations in Mexico City soft soils

Figure 19. Slice model for simplified 3D FEM analysis of piles groups [50].

Important conclusions were established:


a) For the design of piles, it is very conservative to
add the effect of negative skin friction to
accidental loads. Indeed, when a pile is
subjected to the effect of depletion of an initial
pore pressure profile, if a load is applied to the
head of the pile, it behaves similarly to a
preloaded element, i.e. an important part of the
force generated by negative skin friction is
replaced by the surcharge and in some cases
friction can become positive. As set forth by
Fellenius [22], the problem of downdrag forces
on piles is thus mainly a problem of deformation.
This has been taken into account in the 2015
review of Mexico City building code to be
published soon where it is explicitly indicated
that: This friction should be considered only for
the structural revision of piles and for estimating
the movements of the foundation (settlement or
emersion).
b) The depth of the neutral level (Fig. 8) tends to
stabilize as the consolidation process unfolds.
The depth of this level depends significantly on
the initial loading conditions of the pile. The
position of the neutral level depends more on
the spacing between piles than on the
magnitude of pore pressures depletion.
c) The interaction between pile and soil is very
sensitive to variation of the position of the
groundwater level (WT). The transition between
the dry and rainy seasons (raising WT) can
generate the development of two neutral levels
along the pile as was shown by Auvinet and
Hanell [4].
d) For the numerical modeling of long-term
behavior of a central pile which is part of a

e)

f)

g)

h)

i)

j)

group, it is not necessary to use interface


elements at the soil pile contact. This is because
their behavior depends more on the
compressibility of the material than on its shear
strength.
The initial anisotropy of compressible material
influences the behavior of the piles when the
spacing between these elements is relatively
large and when the stress level in the medium is
low, i.e. in those cases in which the shape of the
plastic yield surface is more important than the
position of the failure criterion.
Negative skin friction on point bearing piles due
to depletion of pore pressures, cannot lead to
loss of confinement (effective vertical stress
smaller than the initial value) at the level of the
support stratum, contrary to the concept
proposed by Zeevaert [63].
Qualitatively, rigid inclusions present the same
behavior as piles, however, the former are less
efficient. This is because the external load on
inclusions is transmitted both by the head and
the shaft requiring some deformation of the
distribution layer and the reinforced layer.
The 3D models show that 2D axisymmetric
models (cells representing the tributary volume
of each pile; [13]) are applicable to foundations
formed by a large number of piles.
For the modeling of small groups of piles it is
necessary to use 3D models, since in this case
the edge and corner elements have a major
influence on the overall behavior of the
foundation.
Stiffness of the surface layers and slab or box
foundation is a factor that greatly influence the
behavior of foundations formed by small groups
of piles.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

AUVINET G. et al.

k) The behavior of the corner and edge piles,


submitted to depletion of pore pressures, is
practically independent of the plan dimensions
of the foundation, provided that the separation
between elements is the same.
l) For small groups of piles, stress concentration
at the bottom of the piles due to depletion of
pore pressures, generates a significant bending
of the elements located in the perimeter and, to
a lesser degree, of central piles.
m) Limit shear strength conditions may be
generated by negative friction on the shaft of the
perimeter piles. For central piles, this condition
is generally reached only at the tip and in the
lower part of the shaft where positive skin
friction prevails. Limit conditions for the negative
skin friction on central piles can only be attained
for large spacing between piles and high
piezometric drawdown.
n) Numerical modeling confirm that negative
friction on central piles in consolidating soils
cannot be greater than the apparent increase of
the submerged weight of the mass of soil
surrounding the pile above the neutral level,
induced by seepage forces associated to the
head gradient caused by pore pressure profile
depletion.
o) The analytical model by Resndiz and Auvinet
[46] modified to take into account the
considerations set forth in the preceding
paragraphs gives acceptable results and can
continue to be used as a simple tool for the
analysis of piled foundations in consolidating
soils.

5 CONCLUSIONS
The difficult geotechnical conditions prevailing in the
lake zone of Mexico City, have led to the
development of multiple solutions for foundation of
high-rise buildings in soft soils affected by regional
consolidation. Many valuable contributions to the
analysis of foundations in these conditions have been
made in recent decades. The role of numerical
methods to obtain a better understanding of the
behavior of these foundations has been particularly
significant. New construction methods also have a
strong influence on the deep foundation engineering
practice.
REFERENCES AND BIBLIOGRAPHY
[1] Aguilar, J.M. and Rojas, E., Importantes mejoras
en los dispositivos de control de pilotes, Memoria
de la XVa Reunin Nacional de mecnica de
suelos, San Luis Potos, Sociedad Mexicana de
Mecnica de Suelos, now Sociedad Mexicana de

65

Ingeniera Geotcncia, (SMMS now SMIG), Mxico, 1990.


[2] Aguirre, M., Dispositivo para controlar hundimientos de estructuras piloteadas, Publicacin No
439, Instituto de Ingeniera, UNAM, Mxico, D.F.,
1981.
[3] Auvinet, G., and Daz Mora, C., Programa de
computadora para predecir movimientos verticales de cimentaciones, Publicacin N 438 del Instituto de Ingeniera, UNAM, 71p., junio, Mxico,
1981.
[4] Auvinet, G. and Hanell, J.J. Negative skin friction
on piles in Mexico City clay, Proc. Xth International Conference on Soil Mechanics and Foundation Engineering, 2 (1981), 599-604.
[5] Auvinet, G. and Mendoza, M., Comportamiento de
diversos tipos de cimentaciones en la zona lacustre de la Ciudad de Mxico durante el sismo
del 19 de Septiembre de 1985, Memoria, Simposio "Los sismos de 1985; casos de mecnica de
suelos", SMMS, (1986), 227-240.
[6] Auvinet, G. and Mendoza, M., Consideraciones
respecto al diseo de cimentaciones sobre pilotes de friccin, VIIa Reunin Nacional de Ingeniera Ssmica, (1987), 19-21.
[7] Auvinet, G. and Gutirrez, E., Instrumentacin de
un edificio en proceso de recimentacin, Memoria, Simposio sobre recimentaciones, SMMS,
(1989), 137-148.
[8] Auvinet, G., and Rossa, O., Reliability of Foundations on Soft Soils, Proceedings, Sixth International Conference on Applications of Statistics
and Probability in Civil Engineering, CERRAICASP-6, (1981), 768-775.
[9] Auvinet, G., Pecker, A. and Salencon, J., Seismic
bearing capacity of shallow foundations in Mexico City during the 1985 Michoacn Earthquake,
Proceedings, Eleventh World Conference on
Earthquake Engineering, (CDROM), (1986).
[10] Auvinet, G. and Rodrguez, J.F., Modeling of friction piles in consolidating soils, Proc. Int. Deep
Foundation Congress, ASCE, (2002), 224-235.
[11] Auvinet G. and Rodrguez, J.F., Behavior of endbearing piles in consolidating soils, Proc. Int.
workshop, Foundation Engineering in difficult soft
soil conditions, ISSMGE TC 36, (2002), 133-137.
[12] Auvinet, G and Rodrguez, J.F., Inclusiones rgidas como alternativa de cimentacin en suelos
lacustres de la ciudad de Mxico, Memoria, XXIII
Reunin Nacional de Mecnica de Suelos e Ingeniera de Cimentaciones, Sociedad Mexicana
de Mecnica de Suelos, (2006).
[13] Auvinet, G. and Rodrguez, J.F., Modelling of rigid inclusions in consolidating soils, CD-ROM Proceedings, XIIIth Pan-American Conference on
Soil Mechanics and Geotechnical Engineering,
(2007)

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

66

Deep foundations in Mexico City soft soils

[14] Auvinet, G., Land subsidence in Mexico City,


Proceedings, ISSMGE TC36 workshop, Geotechnical engineering in urban areas affected by
land subsidence; the cases of Mexico City,
Bangkok and other large cities, (2009), 3-11.
[15] Auvinet, G., Soil fracturing induced by land subsidence, in Land subsidence, Associated Hazards and the Role of Natural Resources Development, 339, (2011), 0-26
[16] Auvinet, G. et al., Evaluation of regional subsidence and soil fracturing in Mexico City Valley,
This Conference (2015).
[17] Correa, J.J., The application of negative friction
piles to reduction of settlement, Fifth International
Conference on Soil Mechanics and Foundation
Engineering, (1961).
[18] Correa, J.J., A telescopic type of pile for subsidence conditions, Proc. Specialty session on negative skin friction and settlements of piled foundations, 7th International Conference on Soil
Mechanics and Foundation Engineering, (1969).
[19] Correa, J.J., Estado actual del conocimiento sobre pilotes de control, Memoria de la Reunin
Conjunta Consultores-Constructores Cimentaciones profundas, SMMS, (1980).
[20] Cuevas, J.A., The floating foundation of the new
building for the National Lottery of Mexico: an actual size study of the deformations of a flocculent
structured deep soil, First International Conference on Soil Mechanics and Foundation Engineering, Harvard, (1936).
[21] Ellstein, A., El pilote penetrante o P3, Memoria
de
la
Reunin
Conjunta
ConsultoresConstructores Cimentaciones profundas, SMMS,
(1980).
[22] Fellenius, B.H., Recent advances in the design of
piles for axial loads, dragloads, downdrag and
settlement, ASCE and Port of NY&NJ Seminar,
(1998).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

The use of micropiles technology in soft soil conditions


El uso de tecnologa de micropilotes en condiciones de suelo blando
Federico PAGLIACCI1
1

Sc. in C.E. - Soilmec

ABSTRACT: Micropiles are small diameter structural elements that can be used for deep foundations and underpinning,
soil consolidation and retaining supporting walls.Starting from the first application in 1952, the use of micropiles in
underground engineering has increased considerably thanks to the versatility of the technology and the possibility to use
less cumbersome equipment, resulting in a reduced impact on existing soil and superstructures.This technique is applicable
in soft to firm ground conditions - in loose to medium dense sands and in cohesive soils.This presentation will present an
overview of the various types of micropiles that have been used, the different drilling techniques that can be used according
to soil nature and characteristics, general requirements for types of equipment and tooling needed, and design and
construction recommendations. In addition, practical examples of the technology will be presented as mini case histories.

1 INTRODUCTION
The use of micropiles in underground engineering has
increased considerably thanks to the versatility of the
technology and the possibility to use less
cumbersome equipment, resulting in a reduced
impact on existing soil and superstructures. Usually
micropiles are described as: "Small diameter
structural elements that can be used for deep
foundations and underpinning, soil consolidation,
retaining walls for deep excavations and tunnelling".
A second, more recent definition is the following: a
small-diameter (less than 300 mm), replacement,
drilled pile composed of placed or injected grout, and
having some form of steel reinforcement to resist a
high proportion of the design load. (D. A. Bruce,
1999).

The origin of micropiles is certainly the work of Prof.


Lizzi (1914-2003), who in 1952 registered the patent
of the "root piles", small diameter piles for
underpinning of buildings subject to excessive
settlement (Figure1).
Thanks to the fact that the piles could be built with
very small and lightweight equipment and the
subsequent possibility of operating in confined areas,
the method was broadly used for underpinning historic
buildings like Ponte Vecchio in Florence in 1966 and
stabilising the leaning bell tower in Burano, Italy
(Figure 2).

Figure 1. Typical scheme of Root piles for the consolidation of an ancient monument; the block of soil resulting from the
reinforcement with metal "roots" acquired adequate mechanical properties to withstand the applied loads.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

68

The use of micropiles technology in soft soil conditions

Figure 2. Root piles adopted for the consolidation of the leaning bell tower of Burano, Italy.

Thanks to the fact that the piles could be built with


very small and lightweight equipment and the
subsequent possibility of operating in confined areas,
the method was broadly used for underpinning historic
buildings like Ponte Vecchio in Florence in 1966 and
stabilising the leaning bell tower in Burano, Italy
(Figure.2).
2 MICROPILES CLASSIFICATION
The micropiles construction consists of three main
stages:
a) drilling
b) laying of the reinforcement steel element
c) grouting
The most commonly used method to classify
micropiles is based on the type of selected grout
injection method.
The two main types of microplies can be classified
as follows:
1) Low-pressure grout-injected microplies
2) High-pressure grout-injected micropiles

2.1 Low-pressure grout-injected micropiles (gravity


backfilling)
Micropiles are executed by simply filling the hole after
laying the reinforcement steel element. The hole is
filled from the bottom up using a water/cement mix,
adding sand if needed. If tubular steel reinforcements
are used, the tube itself acts as a tremie pipe through
which the mix is pumped to the bottom of the hole and
gradually fills it up to the surface of the tube, bonding
it to the surrounding soil.

(manchettes). Grouting is a two-step process. The first


step is identical to the one adopted for low-pressure
injected micropiles: the space between the tube and
the hole wall is filled from the bottom with the grout.
After hardening of the initially placed grout, the second
stage of grouting is performed. Through the nonreturn valves, by using a double packer, the grout is
high-pressure injected in a selective way through each
valve; this injection creates a bulb of consolidated soil
that significantly increases the load bearing capacity
of the micropile.
3 DRILLING TECHNIQUES
The techniques used for drilling micropiles are
those usually labelled "small diameter drilling"
techniques. The key advantages of using these
techniques is the possibility of drilling a large variety
of rocks and soils, overcoming pre-existing
foundations and other obstacles, boulders and rocky
layers. Depending on the types of soils the following
drilling techniques can be used: rotation with or
without casing; rotation-percussion by means of
Down-the-Hole Hammer (DTH) with or without casing;
rotation-percussion by means of top hammer without
casing.
In the Table 1 the standard diameter and the
maximum depth are reported, according to the three
above listed drilling methods.
A correlation between the drilling method and the
type of soil, according to ASTM classification, is
reported in Table 2.
The same correlation, based on International
Drilling Company (IDC) (Table. 3.1) is reported in
Table 3.2.

2.2 High-pressure grout injected micropiles


In this case, the micropile tubular reinforcement
bottom section is equipped with non-return valves

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

PAGLIACCI F.

69

Table 1. Micropiles standard diameters and maximum depth function of drilling technique.
Standard diameter
Max. drilling depth
Drilling technique
(mm)
(m)
Rotation with or without casing

80 - 400

50 - 70

Rotation-percussion by means of Downthe-Hole Hammer (DTH) with or without


casing

80-230

60 - 150

Rotation-percussion by means of top


hammer without casing

50-250

15 - 30

Table 2. Correlation between the drilling method and the type of soil, according to ASTM classification.
Type of soil

Clay/Silt

Sand
Fine

Medium

Gravel
Coarse

Fine

Cobbles Boulders

Medium

Drilling Method
Rotation
Top hammer
DTH

Rock classification

Table 3.1. IDC rocks classification.


Type of rock

UCS (MPa)

Soft

Coal - Chalk - Marl - Weathered sandstone

2 - 50

Medium

Tuff - Slate - Dolomite - Limestone - Rhyolite

10 - 100

Hard

Limestone - Sandstone - Rhyolite

50 - 200

Very hard

Basalt - Diorite - Gneiss - Schist - Granite - Conglomerate > 200

Table 3.2. Correlation between the drilling method and the type of rock, according to IDC classification.
Rock classification
Drilling method
1
2
3
4
Rotary (rock drilling bt)
Rotary (tricone)
Rotary (diamond crown)
DTH

3.1 Rotary drilling without casing


This technique is used when there is no need to
support the borehole walls or when the drilling fluid
(water, bentonite, polymer) is capable of supporting
the hole walls. Energy is transferred to the drill bit
through the drill rods that are rotated and pushed by a
rotary drive mounted on the drill rig.
3.2 Rotary drilling with casing
This technique is used when there is the need to
support the borehole walls or when use of drilling fluid

is not allowed by project restrictions. A double rotary


drive can be used: one transfers torque to the string
of inner rods, while the lower one rotates the casing.
3.3 DTH drilling without casing
This technique is used when drilling rock.
Percussion is applied on the bottom of the hole by a
compressed-air driven hammer mounted at the
bottom of the drill string through which compressed air
is driven to the hammer. Once percussion is
completed, air is driven back through the drill bit to
clean all the drilling debris from the hole. Torque and

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

70

The use of micropiles technology in soft soil conditions

down crowd force are transferred by a rotary drive


mounted on the drill rig.
3.4 DTH drilling with casing
This technique is used in fractured rocks or in
granular soils when the risk of collapsing of the
borehole is high and drilling fluids do not guarantee
support to the hole walls. The system (ODEX, TUBEX
or SYMMETRIX depending on each manufacturer's
trade mark) is based on the principle of slightly
enlarging the hole diameter during penetration, to
make it larger than the diameter of the casing which is

driven, without rotating, by the down-the-hole hammer


powered by compressed air. The bit has a reamer that
swings out to enlarge the hole diameter, and
subsequently swings to the minimum diameter
allowing the drill string and the DTH hammer to be
lifted up, leaving the casing temporarily in the hole; the
casing will be extracted once installation of the
reinforcement cage and concreting are completed
(Figure 3).

Figure 3. Scheme of the DTH drilling system and detail of the drilling bit and of the swing reamer.

3.5 TOP HAMMER drilling


This technique is used in cohesionless soils and for
shallow boreholes, to install a casing closed by a
disposable bottom end in the soil. Basically it consists
in "driving" in the soil a tube closed at the bottom. The
percussion energy is supplied by a hydraulic hammer
combined with a rotary head, both placed on top of the
casing.
4 MICROPILE STEEL REINFORCEMENT
A steel hollow pipe, a conventional steel cage (vertical
bars and stirrup), or an H beam profile are generally
adopted as steel reinforcement for low-pressure
grout-injected micropiles.
A hollow steel pipe equipped with non-return valves
has to be used for high-pressure grout-injected
micropiles. Non-return valves are located in the
deepest pile section; in general, 2-3 valves per linear
meter of pile are adopted (Figure 4).

Figure 4. Typical steel reinforcement hollow pipe for highpressure grout-injected micropiles.

5 GROUTING
Micropiles are generally grouted with a cement-based
grout. The adopted ratio between cement and water
range from 1 to 2.
The same injection system used for bored piles is
adopted for low-pressure grout-injected micropiles: an
injection pipe is lowered down to the bottom of the
hole and the grout is then injected until it flows back to
the surface; otherwise the same steel reinforcement
hollow pipe is used.
High-pressure grout-injected micropiles are
injected with a two-step process: injection of cement
sheath and injection of anchorage bulb.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

PAGLIACCI F.

In the first stage, the cement mixture is pumped


from the bottom to fill up the space between the hole
walls and the external surface of the steel tube.
This operation is performed to prevent the injection
mix from leaking out during the second stage that is
carried out under pressure. During the second stage
i.e. the injection of the anchorage bulb the grout is
injected through the non-return valves using a double
packer (Figure. 5) lowered in the hollow tube and
placed in a way that isolates every single valve.
The pressurised mixture being injected breaks the
cement sheath and penetrates the soil.
Once the injection of all the valves is completed, the
tube is washed to allow any subsequent injection if
needed.
As a general rule, injection of each valve is stopped
upon reaching the maximum design injection pressure
and/or the set flow rate.
At the end of the injection operations, the tube is
filled with mixture. This technique is similar to the one
used for some years in soil consolidation, where the
mixture is injected through PVC pipes equipped with
non-return valves to fill the empty space in the soil.

71

In micropiles, the mixture is used to create a sort of


"bulb" around the tube and allow transfer of loads to
the soil.
In the bulb injection stage, which is performed
through each individual valve, the injection pressure,
after the initial peak of about 6 MPa needed to "break"
the sheathing mixture and then penetrate the soil,
must be maintained between 2-3 MPa, with flow rate
values ranging from 10 to 50 l/min.
These pressure and flow rate values are necessary
to let the mixture fill evenly the empty spaces in the
soil and compact the soil around the drilled area thus
creating a true bulb without any breaking (claquage)
that may cause the mixture to infiltrate along the
fracture lines far away from the stem, with little or no
effect.
In order to check the said pressure and flow rate
parameters through each valve, it is necessary to
isolate the valve. This is possible by using a packer
that is lowered inside the tube down to the height of
the valve to be injected. The two packers (upper and
lower) are then stuck to the tube inner walls
(mechanically or hydraulically) so as to isolate the
tube section.

Figure 5. Hydraulic double packer.

6 BEARING CAPACITY DESIGN METHOD.


Low-pressure grout-injected microplies are
performed by reproducing the large diameter bored
piles technique; therefore, the load bearing capacity
has to be evaluated by adopting the same criteria
used for bored piles.
High-pressure grout-injected micropiles are
designed by considering only the bearing strata, i.e.
the level where the bond length is grouted.
The design method of high-pressure injected
micropiles (single stage or multiples stages) is based
mainly on the theory of Bustamante-Doix, 1985.
Micropiles are assumed to be formed by a free
length, where no load is transferred to the soil, and a
bond length Lb where the load-transfer mechanism
develops.
In this section, because of the performed grouting
throughout the valves, the grouted borehole is
enlarged via hydro-fracturing of the grout mass to give
a grout root around the core diameter of the borehole
(Figure 6).

Figure 6. Scheme of High- pressure grout-injected


micropile.

In the bond length, because of the injection of the


fluid through the valves, the hole diameter is enlarged
to the diameter Ds that exceeds the drilled hole
diameter Dd.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

72

The use of micropiles technology in soft soil conditions

Considering that the load bearing capacity of a


high-pressure injected micropile is due only to the
lateral friction along the pile shaft and that therefore
the base load bearing capacity is ignored, the ultimate
load bearing capacity (Qu) can be calculated with the
formulas (1) and (2)
=

(1)

=
(2)
where:

presence of stratified soils with very strong variations


of the stress-strain ratio. The very low layers'
contribution to load bearing capacity can be ignored.
As a general rule, the bulb has to start at a depth of
not less than 5 meters below the soil level with a
length not less than 4 metres.
The recommended safety factors for the working
load, a function of the use of micropile, temporary or
permanent, and of the type of stress, tension or
compression, are reported in the Table 4.
Table 4. Safety factor for the evaluation of the micropile
working load
Safety factor

= average actual diameter of the bulb

= drilled hole diameter

= non-dimensional coefficient, a function of the


nature of soil, injection method and volume of
injected mixture

= bond length

= bulb soil ultimate adhesion

If the soil is stratified, the load bearing capacity of


micropiles can be calculated by summing the effects
of the different layers in which the bulb has been
constructed, using the formula (3):
= 1

(3)

The symbols have the same meaning as above,


while the "i" subscript indicates that the values pertain
to the "i" layer among n layers the bulb runs through.
Care must be paid to using the last formula in the

Scope of work

Tension

Compression

Temporary micropile

1,8

Permanet micropile

2,2

The values of the non-dimensional coefficient, a


function of the nature of soil, injection method and
volume of injected mixture, are reported in the tab. 5.
The table details the minimum amounts of mixture to
be injected. The Vp value pertains to the drilling
volume not including the micropile reinforcement pipe
steel volume.
The bulb soil ultimate adhesion qa, function of the
Nspt value has been evaluated by BustamenteGianeselli for cohesionless and cohesive soils and it
is reported in the Figures 7 and 8.

Table 5. non-dimensional coefficient and Minimum quantity of injected grout for different type of soils.

coefficient
Soil nature

Single stage
grouting

Multiple stage
grouting

Minimum quantity of injected grout


Vi

Gravel

1,3

1,8

1,5 Vp

Sandy gravel

1,2

1,7

1,5 Vp

Gravelly sand

1,2

1,5

1,5 Vp

Sand (rough to fine)

1,1

1,4

1,5 Vp

Silty sand

1,1

1,4

Silt

1,1

1,4

Clay

1,2

1,8

1,5 Vp (Single stage grouting)


2,0 Vp (Multiple stage grouting)
2,0 Vp (Single stage grouting)
3,0 Vp (Multiple stage grouting)
2,0 Vp (Single stage grouting)
3,0 Vp (Multiple stage grouting)

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

PAGLIACCI F.

73

Figure 7. qa values for clay and silt and for simple injection (single stage grouting in tab.5) or repeated injection (multiple
stage grouting in tab.5).

Figure 8. qa values for sand and gravel and for simple injection (single stage grouting in tab.5) or repeated injection
(multiple stage grouting in table.5).

6.1 Numerical example of calculation of load


bearing capacity
As a theoretical example, the working load for a
permanent, vertical compressed, micropile is
calculated. The working load will be calculated for
single stage grouting.
The drilling diameter is assumed equal to 250 mm.

The 12 meter-bond length is assumed to be


performed in a homogeneous layer of sand with Nspt
value equal to 50.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

74

The use of micropiles technology in soft soil conditions

= = 0,35 * 12 * 30,6
= 403,55 ton
Wl = 201,78 ton
7 MAIN APPLICATIONS

Figure 9. Calculation scheme.

The coefficient value for single stage grouting, in


sand, is equal to 1,1.
Therefore, the average actual diameter of the bulb
is equal to:
= = 1,1 * 0,25 = 0,28 m
The bulb soil ultimate adhesion qa, value is
obtained by entering the horizontal axis of figure 8 at
the 50 value of SPT and intercepting the line 2 of the
graph. The corresponding qa value is equal to 0,26
MPa (26,5 tonf/m2), as reported in Figure 10.

Micropiles can be used in several projects that foresee


the construction of foundations with small drill rigs or
whenever the type of soil makes it possible to use
technologies like roto-percussion. The main
applications are: retaining walls; underpinning and
tunnelling.

7.1 Retaining walls


Micropiles can be used successfully to support
boreholes, especially in the presence of coarse
material with boulders or blocks, that is whenever
concrete diaphragm walls cannot be easily built. In
these conditions, since the main stresses come from
a combination of bending and shear, it is better to use
low-pressure grout-injected microplies. Usually,
anchors are installed to reduce the bending moment
along the wall, as reported in Figure 11.
When building an underground parking lot below or
near existing buildings, it is possible to use micropiles
as retaining walls in combination with floor slabs used
as struts, all built with a top-down technique. If
micropiles are constructed in the right positions, they
can indeed become, with an additional reinforcement,
the supporting pillars of the floors as reported in
Figure 12.

Figure 10. Determination of the qa value.

The ultimate bearing capacity is therefore equal to:


=
= 0,28 * 12 * 26,5 = 279,59 ton
The working bearing capacity, for a permanent,
vertical compressed, micropile is therefore equal to:
Wl = Qu /Fs = 279,59/2 = 139,79 ton
Always as an example, if the grouting is performed
as multiple stage grouting, the coefficient value will
be equal to 1,4 and the Db will become 0,35 m.
In this case the qa value in Figure 10 will relate to
the line 1 and therefore equal to 0,3 MPa (30,6
tonf/m2).

Figure 11. Micropiles and anchors for retaining wall.

The ultimate bearing capacity and the working load,


for the same length and drilling diameter of the
micropile will therefore be equal to:

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

PAGLIACCI F.

Figure12. Micropiles as retaining wall and pillars.

7.2 Underpinning
As already mentioned, underpinnings were one of the
earlier applications of micropiles. Several solutions

75

are available, depending on the features of the


existing foundations. In case of old masonry
structures, it is recommended to use a larger number
of micropiles with a relatively small load bearing
capacity to improve distribution and transfer of the
load to the superstructure. The micropiles are usually
connected to the existing foundations by making them
adhere to the masonry work or via a concrete beam
connected the old foundations. Several different
solutions are available for concrete foundations.
In Figure 13, some photos related to the project to
support existing spread foundation plinths during the
construction of a new auditorium under the ground
level of an old industrial facility being renovated. Six
micropiles for each plinth were constructed: four went
through the plinth body and two were placed outside
it. During the under-excavation, a concrete pillar was
vertically cast-in-place for the purpose of enclosing
the metal reinforcement of the micropiles supporting
the superficial foundations.

Figure 13. Example of the use of microplies to support existing spread foundation plinths during the construction of a new
auditorium under the ground level of an old industrial facility being renovated.

7.3 The Pont de Pierre in Bordeaux (France)


The construction of the bridge started in 1811 and
ended in 1821. Stability problems arose from the very
beginning of the construction works. They were due to
the short length of the wooden foundation piles that
did not reach the deep layer of compact gravels
(Figure 14). The problems were further worsened by
significant settlement of the piers caused by the river
stream that removed material from under the bed level

(scouring). In 1985 the settlement, under some of the


piers, had reached 70 cm.
The underpinning works involved only the piers
subject to the most serious settlement forces. It was
also decided that the 16 micropiles, 118 mm diameter,
per pier should bear only 40% of the load weighing on
the pier foundation. The micropiles had to bear a load
of 300 tons each. The intervention stopped the
increasing settlement of the pier.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

76

The use of micropiles technology in soft soil conditions

Figure 14. The Pont de Pierre in Bordeaux (France) and a typical cross section of a pier.

Figure 15. 16 micropiles per pier have been executed from the bridge deck. The micropiles had to bear a load of 300 tons
each.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Pore pressure build-up due to pile driving in clayey deposits


Desarrollo de presin de poro debido al hincado de pilotes en depsitos arcillosos
Manuel J. MENDOZA 1, Miguel RUFIAR 2, Enrique IBARRA 2 and Marcos OROZCO 2
1

Research Professor, Instituto de Ingeniera, Universidad Nacional Autnoma de Mxico


Former graduate student, Instituto de Ingeniera, Universidad Nacional Autnoma de Mxico

ABSTRACT: Experimental signals are depicted on a) the pore water pressure build-up and the effective stresses on the
shaft of an instrumented pile model that are recorded in the laboratory during pile driving into reconstituted marine clayey
soil and b) pore water pressure build-up, which was measured in lacustrine clay from Mexico City with a given distance
between the piles during pile driving at a site where a friction pile-box foundation was built. Experimental results are
analyzed in terms of different theoretical solutions and highlight the similarities and differences in the comparison
between the predictions and data recorded in the laboratory and field.

1 INTRODUCTION
There is currently significant interest in understanding
the change in pore pressure in the clayey subsoil that
surrounds a field of piles because its magnitude has
considerable influence in the pile fields axial load
capacity. This study attempts to describe the change
in pore pressure during the pile-driving process and
that immediately after the completion of pile driving.
When a pile is driven by hammering, pore
pressures are induced in the surrounding soil that are
so high that they surpass the pre-impact in situ
effective vertical stress; this significantly affects the
penetration of the pile and leads to an immediate
load bearing capacity that is near zero. During the
process of pile driving in the soft clayey soil of Mexico
City, it is thus necessary to bind the piles with steel
wires and straps because after only a few controlled
hammer impacts, the pile easily penetrates the soil
by several meters. Due to the subsequent dissipation
of the pore pressure in the water, and the resulting
gain in shear strength along with other phenomena
that occur around the shaft, which are discussed in
this article, the piles gradually achieve the capacity to
withstand a given workload.
This study reports on the pore pressure and the
change in the effective stresses measured at the
lateral face of a scaleddown model pile while it is
driven into the soil by impact driving. These
measurements were performed under controlled
conditions in the laboratory using a reconstituted
clayey soil of marine origin that was obtained from
the sea bottom at the Sonda de Campeche in the
Gulf of Mexico. The reconstituted soil was made

since a suspension condition, which was contained


within a vessel with a diameter of nearly one meter.
In addition, this study shows the pore pressure
measurements recorded in the saturated mass of a
clayey soil under an instrumented friction pilebox
foundation at the instant when the piles were driven
to a certain depth. The subsequent change in the
pore pressure is also shown. This case corresponds
to a foundation composed of a concrete box and
friction piles, which is typical of the foundations that
are frequently built in the lacustrine zone of Mexico
City.
The experimental observations in this study are
analyzed using different theories to predict the
increase in pore pressure that occurs during pile
driving, and its dissipation over time; in particular, the
similarities and differences between the laboratory
and field observations, and the theoretical predictions
are highlighted.
2 BACKGROUND AND OBJECTIVES
In the geotechnical literature, there are various semiempirical and analytical methods for predicting the
increase in pore pressure that occurs during pile
driving and its dissipation with time, which leads to a
resulting increase in resistance. Bogard and Matlock
(1990) proposed correlations for predicting the time
required to reach different resistance levels based
only on local consolidation mechanisms. DAppolonia
and Lambe (1971) presented a method for the
calculation of the pressure increase and its
dissipation over time, while Vesic (1972) obtained an
expression for estimating the increase in pore
pressure near the shaft.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

78

Pore pressure build-up due to pile driving in clayey deposits

The primary objective of this study is to succinctly


review these methods and carry out a comparison
between their predictions and two experimental
observations of piles driven under controlled
conditions. One of these experiments was performed
using a small-scale model studied in the laboratory;
the other experiment was performed using real piles
in the field. The first measurements are part of a
research project sponsored by the Mexican
Petroleum Institute (Instituto Mexicano del Petrleo),
testing an instrumented model pile. The results of
this experiment have already been presented in
previous papers (Mendoza et al., 2000a and Luna,
2002). The pore pressure and total pressure
measurements on the model pile shaft help to define
the effective stress state along the pile and its
variation over time.
Conversely, pore pressure measurements of the
soil mass during the driving of 77 piles in an actual
foundation, corresponding to Support No. 6 of the
Impulsora overpass in northeastern Mexico City are
also presented. The details of the stratigraphic
conditions,
foundation
characteristics,
and
instrumentation have been described in previous
papers (Mendoza et al., 1996; Mendoza et al.,
1998a).

a function of the distance from the pile shaft and


dissipates quickly over time. Poulos and Davis (1980)
reported measurements of the increase in pore
pressure u that were normalized with regard to the
pre-driving in situ effective vertical stress vo and
presented them as a function of the radial distance d
from a driven pile of radius r (Figure 1).

3 THEORETICAL SOLUTIONS FOR THE


EFFECTS OF INSTALLING PILES IN CLAY
The method used to install piles affects the loadstrain behaviour of the soil-pile system due to
changes in the initial state of the soil. The effects of
piles driving into clay have been classified into four
broad categories by De Mello (1969):
Remoulding or partial alteration of the soil
structure near the pile;
Changes of the stress state of the soil in the
vicinity of the pile;
Increase in pore pressure due to driving and its
dissipation near the pile; and
Aging phenomena.
This article focuses only on the pore pressure and
its change over time. Some aspects related to aging
have been presented in previous studies published
by this group (Cruz, 2003).
3.1 Increase in pore pressure due to pile driving
Several authors have reported measurements of
excessive pore pressure in soil due to pile driving,
including DAppolonia and Lambe (1971) and
Randolph et al. (1979). Some researchers have
recorded the pore pressure along the pile shaft and
found that it can reach or surpass the magnitude of
the vertical effective stress by up to three times.
However, this pressure excess decreases quickly as

Figure 1. Summary of the pore pressure measurements


around a driven pile (Poulos and Davis, 1980).

Beyond a ratio d/r of 4 in non-sensitive clays and


of approximately 8 for sensitive clays, the pore
pressure is shown to decrease rapidly as distance
increases. Beyond d/r = 30, the excess pore
pressure is effectively zero.
The data presented by Airhart et al. (1969)
suggest that pore pressures are higher near the end
of the pile (i.e., approximately three to four times the
in situ effective vertical stress). Both phenomena are
clearly observed in the measurements of the model
pile presented and described below. Figure 1
includes the point corresponding to the measurement
near the tip and at the interface (d/r=1) of the model
pile after driving
3.2 Estimation of the increase in pore pressure at
the pile shaft
To predict the increase in pore pressure at the pile
shaft due to driving, DAppolonia and Lambe (1971)
proposed the following expression (1).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MENDOZA M. et al.

u m
2s
1 K 0 u A f
(1)
vo
vo

where: u m = maximum increase in pore pressure;

K0= earth pressure coefficient; su = undrained shear


strength; Af = Skempton pore pressure coefficient at
failure, and vo = vertical effective stress in the soil.
Comparisons reported by Lo and Stermac (1965)
and Lo (1968) between Equation (1) and measured
pile pore pressures after pile driving show an overall
good agreement.
Based on the expansion theory of cylindrical
cavities, Vesic (1972) considered the stress-strain
changes in undrained conditions and evaluated the
increase in pore pressure at the shaft to be:

R p
r su

where:

f 0.707(3 Af 1)

u 0.817 f 2 Ln

Af

(2)

q 1 sen

su
2 sen

where: q = initial isotropic effective stress; =


internal friction angle, in terms of effective stresses.

Rp I r
where:

Ir

1
r
cos

that the dissipation occurs only radially in a process


described by the following partial differential
equation.

2u 1 u
u
ch 2
t
d d
d

(6)

where: ch = consolidation coefficient in two


dimensions, for horizontal drainage; u = excess pore
pressure.
From a practical perspective, solutions to Eq. (6),
such as that shown in Figure 2 (Poulos and Davis,
1980), are used to estimate the time that must pass
after driving before a load test can be performed.
A rigorous analysis of the increase in pore
pressure and the subsequent consolidation around a
pile driven into clay was performed by Wroth (1979).
During the driving process, the pile is modelled as
the formation of a large cylindrical cavity. The
changes in the stress and pore pressure were
obtained via finite element analysis, where the Camclay model was incorporated. In these studies, it was
concluded that the effective and total stresses
adjacent to the pile immediately after driving may be
directly related to the original shear strength of the
soil and are essentially independent from the preconsolidation ratio.

(3)

R p = radius of the plastic zone.

E
2(1 ) su

where:

79

(4)

I r = rigidity index; E = undrained soil elastic

modulus; = Poisson ratio of the soil


Conversely, Randolph et al. (1979) suggested that
the excess pore pressure due to pile driving can be
estimated by the following expression:

u 4su 'vo

Figure 2. Dissipation of the pore pressure near a pile


(Poulos and Davis, 1980).

(5)

where: vo = the change in the effective stress due


to the remoulding of the soil. In normal or marginally
overconsolidated clays, vo can reach negative
values; for sensitive clays, it reaches two to three
times the value of su.
3.3 Dissipation of the excess pore pressure
Soderberg (1962) proposed a relatively simple
solution for predicting the dissipation of the increase
in pore pressure around driven piles; he assumed

4 MEASUREMENTS ON A MODEL PILE


Experimental measurements were performed on an
instrumented model pile that was 90 cm in length and
2.64 cm in diameter. The descriptions of the model
pile, its instrumentation, the data acquisition systems
used, and the applied loads as well as the soil where
it was driven have been reported in previous studies
(Mendoza et al., 1998 and Mendoza et al., 2000);
only a brief description is presented here.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

80

Pore pressure build-up due to pile driving in clayey deposits

The model features four axial-load and three


bending transducers (CF) along its body, as shown in
Figure 3. In addition, there are four total-pressure
(PT) and four pore-pressure (PP) transducers along
its lateral face at different heights.

Figure 4. Experimental process with the model pile.

The soil inside the O-97-5 oedometer used in this


study corresponds to reconstituted marine clay from
the Campeche Sound (Ibarra, 2002) having an overconsolidated condition (OCR=2) with a uniform
pressure of 75 kPa along its surface. This pressure is
kept constant throughout the test including during the
driving process.

During driving, all model pile sensors were


monitored and this activity was kept for nearly 24
hours after the driving process.
The system set for applying loads to the pile
consisted of a reaction frame, an electro pneumatic
servo knob, a pneumatic actuator for the axial load, a
load cell, and an LVDT linear displacement
transducer. In addition, it featured a signal
conditioning module, an analog/digital/analog (A/D/A)
computer card, a computer interface, a servo
amplifier for the servo knob control, and the software
that controlled all of these components. The tests
were performed (Rufiar, 2009) using a program that
includes static and axial dynamic tests.

4.1 Brief description of the experiments

4.2 Experimental results with the instrumented pile

The experimental procedure of this study consisted


of five stages, which are shown schematically in
Figure 4 and are described below.

The measurements showing the increase in pore


pressure due to pile driving from two tests, which are
denoted A1 and A3, are presented below.

The driving of the pile was performed with the aid


of a steel guide that directs the model pile while it is
impact-driven. The impacts were provided via a
stainless steel mass that had a mass of 2.98 kg
(29.24 N). The mass was allowed to fall freely from a
fixed height of 0.20 m. In general, driving of the
model piles required 330 impacts.

Figure 5 shows the driving data, where the number


of impacts required to drive the pile by 10 cm are
reported.

Dimensions in mm

Figure 3. Schematic of the instrumented model pile.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MENDOZA M. et al.

81

an asymptotic trend, as shown in Figure 6, which


confirmed the observations of Ahrart et al. (1969).
A similar behaviour was observed in test A3, as
shown in Figure 7, where a maximum increase on
the order of 110 kPa was observed for cell PP3,
which is located in the middle of the pile, which
dissipated in the same manner as shown in test A1.

Figure 5. Records of pile-driving tests A1 and A2.

It is shown that 13 to 18 impacts were necessary


to drive the pile by the initial 10 cm; conversely, for
the 10-cm section between 0.5 m and 0.6 m, 75 to 78
impacts were required. This result clearly shows the
increase in resistance of the pile shaft as it is driven
into the clayey soil.
Figure 6 shows the pore pressure that was
measured at the shaft during and after the driving
operation. It is shown that the pile driving generates
pore pressures that increase with depth.

Figure 7. Variation in pore pressure, measured at the shaft


of the model pile during and after driving. Test A3.

Certain man oeuvres were required in both tests


that required temporarily stopping the driving a few
minutes after the test had begun. For that reason, an
immediate dissipation of the pore pressure occurred,
as shown in the interval between 1000 and 1500
seconds in Figure 6. Because the pile driving was
resumed, the pore pressure began to increase again.
Figure 8 shows the normalized data from test A1 as
a function of time; the normalization was performed
with respect to the in situ effective vertical stress
'v0=v0-u0. It is shown that the installation of the pile
causes a significant increase in pore pressure. A few
seconds after driving began, an increase in water
pressure u2.28'v0 was measured in cell PP4 near
the end of the pile. In cell PP2, which is near the
head of the pile, the lowest increase in pore pressure
was detected (i.e., u0.88'v0).

Figure 6. Variation in pore pressure, measured at the


model pile shaft during and after driving. Test A1.

The magnitude of the maximum increase in the


pore pressure at cell PP4 was 170 kPa; for cell PP3,
it was 140 kPa, and for cell PP2, it was only 60 kPa
near the pile head. Thus, the u /vo ratio varied
from 2.5 to 0.9. These pressures were found to
dissipate within approximately 20 hours and followed
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

82

Pore pressure build-up due to pile driving in clayey deposits

the subsoil at different depths in different


perforations; two were placed ex professo in the clay
strata. The piezometers produced high-quality and
rapid responses; as a result, the pore pressure of the
site was immediately known. Three piezometers
were of the resistive type with strain gauges (SG),
and the remaining piezometers were of the vibratingwire type (VW); all were calibrated meticulously
before experimentation. The ZD-2 and ZD-3
piezometers were driven directly into the clayey strata
without a sand pocket; the others were placed in
permeable strata.
Table 1. Piezometer installation depths and type of
surrounding
soil.
_____________________________________________________

Figure 8. Variation in pore pressure, measured at the


model pile shaft after installation.

5 MEASUREMENTS IN A PROTOTYPE
FOUNDATION

5.1 General description


The foundation of support No. 6 of the Impulsora
overpass (Mendoza et al., 1996; Mendoza et al.,
1996; Mendoza, 2004), is the studied case history. It
consists of a foundation box and 77 friction piles. The
piles have a 50-cm square cross section and were
driven to a depth of 30 m. The site is located in Zone
III-Lago Virgen, which ensures that the piles are
embedded in very soft soil of the First Upper Clay
Formation (UCF); the First Hard Layer is at a depth
of 33 m.
Before pile driving began, a perforation with a
diameter of 0.50 m was made for the initial 2 or 3 m
of depth, which corresponded to the surface crust
and artificial previous fills. The pile driving was
performed with a Delmag No. 33 hammer, although it
should be made clear that the energy delivered by
this hammer is significantly larger than that required
for this driving process. The soil resistance was
broken with only two or three impacts, or even by its
own weight, inducing considerably displacements to
the piles. For this reason, they had to be fastened
with steel cables to maintain control of their depth.
5.2 Instrumentation and monitoring system
Some days before beginning the pile driving
operations, six electric piezometers were installed in

Piezometer
Type
Depth, in m
Soil type
_____________________________________________________
ZD-1
SG
7.50
Sandy stratum
ZD-2
SG
10.20
Clayey stratum
ZE-1
VW
24.00
Sandy stratum
ZD-3
SG
27.00
Clayey stratum
ZE-2
VW
34.00
First hard layer
ZE-3
VW
52.00
Deep deposits
_____________________________________________________

The piezometers were placed around a central


nucleus of the foundation, as shown in Figure 9.
Different to the pore-pressure measurements, which
were performed at the soil-shaft interface in the
model piles, the measurements in this case history
were performed in the central portion of the
foundation where there were no piles The piles were
concentrated toward the foundation edges, precisely
under the foundation beams to achieve a better
foundation performance during seismic events. Thus,
the pore-pressure measurements were performed in
the soil mass under the foundation at a distance of
approximately 5 m from the surrounding piles.

Figure 9. Layout of the instrumented foundation

The SG resistive transducers recorded dynamic


variations during the instant of pile driving using
digital gauges that operated at a sample rate of 250

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MENDOZA M. et al.

83

samples per second. The measurements reported


here for the vibrating wire sensors were done with a
portable frequency gauge, and correspond to static
monitoring only.
5.3 Results
The digital measurements of the pore pressures from
the SG transducers detected variations in real time
as any of the surrounding piles were driven.
However, their amplitudes were reduced due to the
distance between the pile and the piezometer. For
this reason, more emphasis was placed on the
medium-term monitoring of the pore pressure in the
subsoil under the central part of the foundation,
which originated from the full pile field rather than
from an individual pile.
Figure 10 shows the pore pressure measured at
six different depths under the foundation and
includes the measurements during driving and days
after driving to demonstrate the change in pore
pressure over the subsequent four weeks. The
increase in pore pressure clearly manifested in the
UCF, resulting from the strong distortions in the
subsoil due to pile driving. Because no pre-boring
was practiced, the presence of the piles generates a
volumetric displacement of approximately 500 m 3;
which was reflected in a land surface expansion of up
to 11.5 cm.
The increase in pore pressure due to pile driving is
shown to be highest at depths near the pile tips and
is independent of whether the piezometer is lodged in
a sandy or in a clayey layer. Note that the curves
describing the pressure change in the piezometers
ZE-1 and ZD-3 located at depths of 24 m and 27 m,
respectively, are parallel. The ZD-3 piezometer
recorded an increase of 21% over the previous
hydraulic pressure; the pore pressures in the zones
closer to the piles must be higher than this value. The
piezometers were separated by a distance of
approximately ten times the length of their side. The
increases in pore pressure measured by the other
piezometers (ZD-1 and ZD-2) lodged in the UCF
were lower, which can be interpreted as a result of
smaller distortions in the shallower portions of the
ground, and in the more permeable zones, where
these increases in pressure dissipate more easily.
Although the pile-driving process clearly had an
effect on the UCF soil in terms of the variation of its
pore pressure, it did not affect the pore pressure in
the First Hard Layer or in the Deep Deposits, thus
reflecting the high permeability of their soils.
Conversely, the speed with which the excess pore
pressure caused by driving dissipates with respect to
the initial in situ condition calls ones attention in
Figure 10.

Figure 10. Measurements of pore pressure during and


after pile driving at Support 6 of the Impulsora overpass.

6 COMPARISONS BETWEEN THE


MEASUREMENTS AND PREDICTIONS
Some of the solutions discussed in Section 3 were
applied for comparison to the measured values from
both the laboratory tests with the pile model and the
field experiment using friction piles for Support No. 6
of the Impulsora overpass. The results are
summarized in Table 2.
Table 2. Comparisons between measured and predicted
values
according with various authors.
_____________________________________________________
Measured

DAppolonia
Cavity
and Lambe
expansion theory
kPa
kPa
kPa
_____________________________________________________
Case

Model pile 140


223
170
Impulsora
overpass
piles
48
71
52
_____________________________________________________

In both the laboratory and field experiments, an


overestimation of the increase in pore pressure due
to driving is shown in the theoretical predictions.
Karlsrud and Haugen (1985) reached a similar
conclusion when comparing their results with other
experiments. For the model pile case at the level-3
sensor, this overestimation is approximately 20%
when considering the cavity expansion theory
solution. With this solution, a prediction is obtained
that agrees with the measurements from the
Impulsora bridge piles. The solution proposed by
DAppolonia and Lambe produces results that are
approximately 50% higher than those measured in
both the laboratory and the field tests.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

84

Pore pressure build-up due to pile driving in clayey deposits

7 CONCLUSIONS
It has been possible to verify through laboratory
measurements and field observations that driving a
pile using impacts in the absence of preboring
generates high water pressures at the soil-shaft
interface. In laboratory tests, pore pressures at the
shaft were shown to reach values that were 2.5 times
the pre-driving vertical effective stress; this increment
shows that at the moment of impact, the effective
stresses on the pile shaft will be lost transiently along
with the shear strength of the soil. As a
consequence, the pile can penetrate easily into soft
clayey soil. It is common practice in the virgin region
of Mexico Citys Zone III to bind piles to prevent them
from sinking out of control after a few hammer
impacts. The dissipation of such pore pressure after
pile driving triggers a local consolidation process.
This phenomenon leads to an adjustment of the
effective stresses around the pile and an increase in
the soils thixotropy and thus determines the load
capacity generated in a friction pile over time. Thus,
after up to a few months, the load capacity can
withstand operational loads.
Several theoretical solutions have been discussed
for calculating the pore pressure that is generated
around a pile after being driven into clayey soil.
Among the theories tested, the cavity expansion
theory provided the estimation that best agrees with
the measurements, both in the laboratory and in the
field.
REFERENCES
Cruz, E. (2003). Static and dynamic test on pile
models. Master Thesis, Divisin de Estudios de
Posgrado, Facultad de Ingeniera, UNAM, Mxico (in
Spanish).
DAppolonia, D. J. and Lambe, T. W. (1971).
Performance of four foundations on end-bearing
piles. J.S.M.F.D., ASCE, vol. 97, SM1, pp.77-93.
De Mello, V. F. B. (1969). Foundations of
buildings on clay. State of the Art Report: 49-136.
Proc. 7th Int. Conf. S.M. & F.E., Mxico.
Ibarra, E. (2002). Reconstitution of a marine
clayey soil in an oedometer for pile models testing.
Master Thesis, Divisin de Estudios de Posgrado,
Facultad de Ingeniera, UNAM, Mxico (in Spanish).
Karlsrud, K., and Haugen, T. (1985). Axial static
capacity of steel model piles in overconsolidated
clay. Publication No. 163. Norwegian Geotechnical
Institute, Noruega, Oslo.
Luna, O. J. (2002). Design, construction and
operation of friction pile models under static and
cyclic loading, Master Thesis, Divisin de Estudios
de Posgrado, Facultad de Ingeniera, UNAM, Mxico
(in Spanish).

Lo, K. Y. and Stermac, A. G. (1965). Induced pore


pressures during pile driving operations. Proc. 6th
Int. Conf. S.M. and F. E., vol 2, pp 285-290.
Mendoza, M. J., Romo, M. P., Barrera, P.,
Olivares, A., Rojas, E., Snchez, J., Luna, O. J. and
Valle, C. (1998). On experimental study of friction
pile models for offshore platforms, Proc. XIX Nat.
Meeting on Soil Mechanics, SMMS, Puebla, pp 303312.
Mendoza, M. J., Romo, M. P., Domnguez, L.,
Orozco, M., Noriega, I. and Velasco, J. (1996).
Instrumentation and behavior of a piled-box
foundations in Mexico city, during its construction and
initial operation, Proc. XVIII Nat. Meeting on Soil
Mechanics, SMMS, Morelia, pp 143-159.
Mendoza, M. J., Romo, M. P., Orozco, M.,
Domnguez, L., Velasco, J. M. and Noriega, I. (1998).
Loads on piles, contact and pore water pressures
induced by earthquake in a piled-box foundation in
Mexico City, Proc. XIX Nat. Meeting on Soil
Mechanics, SMMS, Puebla, Vol. 1: 358-367.
Mendoza, M. J., Luna, O. J., Ibarra, E., Olivares,
A., Barrera, P. (2000). Instrumentation of a smallscale pile model: design and manufacturing, Proc.
XX Nat. Meeting on Soil Mechanics, SMMS, Oaxaca,
pp 321-328.
Mendoza, M. J. (2004). Behavior of a piled-box
foundation in Mexico City, under static and seismic
loading, Doctoral Thesis, Divisin de Estudios de
Posgrado, Facultad de Ingeniera, UNAM, Mxico (in
Spanish).
Poulos, H. G. and Davis, E. H. (1980). Pile
Foundation Analysis and Design, John Wiley & Sons,
New York.
Randolph, M. F. and Wroth, C. P. (1979). Driven
piles in clay effects of installation and subsequent
consolidation-, Gotechnique 29 No. 4, pp 361-393.
Rufiar, M. (2009). Behavior of instrumented pile
models in marine clayey soils under axial static
loading. Master Thesis, Seccin de Estudios de
Posgrados e Investigacin, ESIA-IPN, Mexico (in
Spanish).
Soderberg, L. O. (1962). Consolidation theory
applied
to
foundation
pile
time
effects.
Gotechnique, Vol. 12, pp 217.
Vesi, A. S. (1972). Expansion of cavities in
infinite soil mass, Journal of the Soil Mechanics and
Foundations Division, ASCE, Vol. 98, SM3, Proc.
Paper 8790, March, pp 65-290.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Geotechnical design of the foundation for an office building located at the


transition zone
Diseo geotcnico de la cimentacin para un edificio de oficinas localizado en la zona de transicin
Fernando ARENAS1 & Alberto CUEVAS 1
1Ingenieros

Cuevas Asociados S. C.

2.3 Laboratory tests.

1 . INTRODUCTION
Reference is made to an office building that has been
contemplated to be built at Insurgentes Sur Avenue,
between streets Eje 5 (Eugenia) and Eje 4 (Xola),
Colonia Npoles, in the Transition Zone of Mexico
City. The structure has been designed to have seven
basement floors, ground floor, nine stories and
terraced roof with a concrete frame design. The
excavation to accommodate the basement floors will
reach a depth of 24.0 m below sidewalk elevation.
2 2. GEOTECHNICAL INFORMATION AVAILABLE
2.1 Geotechnical zoning
The site under study is located at the so-called
Transition Zone (Ref. 1) being characterized by
stratigraphic discontinuities produced by crossed
alluvial deposits; their frequency and distribution
depends on the closeness to old gorges in the
western hills. These materials are underlain by clay
strata covering typical deposits of the Hill Zone.
2.2 Field works
To be able to define the local stratigraphy, two
combined borings were drilled, alternating the
techniques of electric cone probing with the Standard
Penetration Test to depths of 40.0 and 45.0 m,
respectively; a boring with selected sampling was
also advanced to recover undisturbed specimens
using the technique of Shelby-type thin walled pipe
samplers. In addition, to define the conditions of the
pore-water pressure a piezometric station was
installed incorporating the following instruments: a
tell-tale pipe driven to a depth of 6.0 m and three
Casagrande-type open piezometers installed at depths
of 15.2, 20.5 and 25.2 m, respectively.

Properly protected and identified samples were


moved to our laboratory to determine their index
properties: natural water content, visual and manual
soil classification, grain size distribution, percentage
of fines and consistency limits. The undisturbed
samples were subjected to unconsolidated undrained
triaxial compression tests and to one-dimensional
consolidation tests.
2.4 Stratigraphic interpretation.
Based on the field works and on the laboratory test
results the following stratigraphy was defined (Fig. 1).
- From 0.0 to 0.2 m, plain concrete slab.
- From 0.2 to 1.1 m, fill constituted by fine and
coarse sand of pumice and andesite origin, with
brick fragments.
- From 1.1 to 3.5 m, superficial crust integrated
by silty clay with fine, medium and coarse sharp
sand, of andesite and quartz origin, with stiff to
very stiff consistency.
- From 3.5 to 6.0 m, sandy silt with stiff to hard
consistency with some lenses of hard
consistency.
- From 6.0 to 8.6 m, dense pumice-type fine to
medium silty sand.
- From 8.6 to 17.6 m interstratifications of organic
silt with stiff to hard consistency and highly
plastic silt with fine to medium sand of quartz
and pumice origin and slightly andesitic origin
with stiff to hard consistency, and highly plastic
silt, with quartz and pumice and some andesitic
origin and stiff to hard consistency with
interbedded fine to medium pumice-type semidense and dense silty sand lenses.
- From 17.6 to 28.5 m on the average, very fine,
fine and medium dense to very dense silty
sands of andesite, and pumice origin.
- From 28.5 to 31.0 m on the average, andesitetype, sharp medium and fine very dense gravel

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

86

Geotechnical design of the foundation for an office building located at the transition zone

with andesite-type coarse, medium and fine


sand and some silt.
- From 31.0 to 33.0 m on the average, fine,
medium and coarse silt with hard to very hard
consistency with fine, medium and coarse
andesite and quartz sand.
- From 33.0 to 34.4 m, dense fine to medium silty
sand of quartz, andesite and pumice origin and
subangular particles.
- From 34.4 to 45.0 m, very dense fine, medium
and coarse silty sand of quartz and andesite
origin and sharp particles, with some sharp
andesite coarse gravel.

3.4 Frictional bearing capacity.


The admissible frictional load bearing capacity was
calculated by using the following expression:
Q fa

mi P l i tg i
FS

(2)

where: Qfa, admissible frictional bearing capacity, t;


P, perimeter of drilled shaft, m; mi, mean effective
stress at strata of interest, t/m2; li, length of drilled
shaft at strata of interest, m; angle of internal
friction of strata of interest; FS, safety factors equal
to 2 and 1.7 for static and dynamic conditions,
respectively.

2.5 Hydraulic conditions.


Based on monitoring of the piezometer station
installed it was found that the telltale pipe registered
a superficial water level beyond a depth of 4.4 m,
whereas piezometers at depths of 15.2, 20.5 and
25.2 m, respectively are dry. It was determined that
perched water exists.

3.5 Total admissible load bearing capacity.


The total admissible load bearing capacity is
determined as follows:

QT Qfa Qpa

(3)

where: QT, total load bearing capacity under


compression, t.

3 FOUNDATION ANALYSIS
3.6 Tensile bearing capacity.
3.1 Seismic coefficient.
The seismic coefficient for structural design was
assumed equal to 0.32 (Ref. 2) corresponding to the
Transition Zone.
3.2 Foundation solution.
Based on the stratigraphic information and on the
characteristics of the structure, the foundation
solution consists in drilled shaft foundations resting at
a depth of 35.0 meters.
3.3 Point bearing capacity.
The point bearing capacity of the drilled shaft
foundations was calculated by means of the following
equation (Ref. 3):
Q pa

o N q Ap
FS

he tensile bearing capacity of the drilled shaft


foundations is equal to the admissible frictional
resistance plus the weight of the element.
3.7 Settlements.
The settlements to be expected as a result of load
transmitted to the mass of soil were calculated using
the following expression (Ref. 3):

(Q pa 0.66 Q fa ) L
10 A p Ec

0.36 Q pa D
10 A p Es

(4)

where: , settlement, cm; D, shaft diameter, cm; L,


shaft length, cm; Ap, cross sectional area, m2; Ec,
modulus of elasticity of concrete for fc= 250 kg/cm2,
221,359 kg/cm2; Es, modulus of elasticity of
foundation soil, 1500 kg/cm2.
3.8 Modulus of vertical reaction.

(1)

where: Qpa, admissible point bearing capacity, t; o,


effective stress at foundation elevation, 23.5 t/m2, Nq,
bearing capacity factor proposed by Berezantsev for
an angle of internal friction = 37o, 100; Ap, area at
tip of drilled shaft, m2; and FS, safety factor equal to
3 and 2 for static and dynamic conditions,
respectively.

The modulus of vertical reaction was calculated by


using the following expression:
kv

QT

(5)

where: kv, modulus of vertical reaction, t/cm; Qt, net


static bearing capacity, t; settlement, cm.
Table 1 summarizes the conditions for different
shaft diameters.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

ARENAS F. et al.

87

N.P.T +43.60
N.P.T +42.60

Nivel

Nivel

Nivel

Nivel

Nivel

Nivel

Nivel

Nivel

N.P.T +38.40

N.P.T +34.20

N.P.T +30.00

N.P.T +25.80

N.P.T +21.60

N.P.T +17.40

N.P.T +13.20

N.P.T +9.00
Estacionamiento 1

N.P.T +4.50

SM-2
N.P.T +0.00

Planta Comercial

SM-1
0

Estacionamiento S - 1

(CH) Arcilla caf olivo de consistencia rgida.

Estacionamiento S - 1

Estacionamiento S - 2

N.P.T -4.71
5

(MH) Limo caf con arena fina pumtica y


cuarzosa
de consistencia
rgida.
Estacionamiento
S-3

Estacionamiento S - 2

N.P.T -7.85

Ceniza volcnica muy compacta.


Estacionamiento S - 3

10

N.P.T -11.00

Estacionamiento S - 4

10

Estacionamiento S - 4

N.P.T -14.14
15

Estacionamiento S - 5

1
5

Estacionamiento S - 5

Estacionamiento S - 6

10

15

N.P.T -17.28

1
2
3
4

11

Estacionamiento S - 6

Estacionamiento S - 7

N.P.T -20.42

20

20
Estacionamiento S - 7

12

N.P.T -23.57
25

25

30

30

35

35

40

40

13

14

15
16
0 10 20 30 40 50
NMERO DE GOLPES

18

17

(GP) Gravas muy compactas gris andesticas angulosas con


arena fina a gruesa andestica.

45
0 10 20 30 40 50
NMERO DE GOLPES

1 (SM) Arena pumtica compacta.


2 (SM) Arena pumtica semicompacta.
3 (OH) Limo orgnico de color negro.
4 (MH) Limo arenoso gris verdoso de consistencia
dura.
5 (OH) Limo orgnico caf muy rgido.
6 (MH) Limo arenoso gris verdoso muy rgido.
7 (SM) Arena pumta compacta con limo gris
verdoso.
8 (MH) Limo gris verdoso muy rgido.

9 (SM) Arena limosa muy compacta gris con


gravas andesticas subredondeadas.
10 (MH) Limo gris verdoso de consistencia
semirgida.
11 (SM) Arena limosa muy compacta gris verdoso
con grumos andesticos redondeados.
12 (SM) Arena limosa muy compacta caf verdoso
(arena fina a media, andestica cuarzosa,
subredondeada).
13 (SM) Arena limosa muy compacta caf rojizo con
gravas (arena fina a gruesa; gravas andesticas
subangulosas).

14 (SM) Arena limosa muy compacta caf


amarillento (arena fina a gruesa andestica,
pumtica, cuarzosa subangulosa).
15 (SM) Arena limosa muy compacta caf rojizo,
fina a gruesa andestica, pumtica con
intercalaciones de limo arenoso caf rojizo.
16 (SM) Arena muy compacta caf verdoso fina a
gruesa andestica subangulosa.
17 (SM) Arena limosa muy compacta caf rojizo.
18 (SM) Arena muy compacta caf verdoso fina a
gruesa andestica subangulosa.

Figure 1. Stratigraphic interpretation.

Table 1. Admissible load bearing capacity versus shaft diameter.


Dimeter
m

Static

0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00

110.5
147.4
184.2
221.0
257.9
294.7
331.6
368.4

Qf a , t
Dynamic
130.1
173.4
216.8
260.2
303.5
346.9
390.2
433.6

Qpa , t
Static

Dynamic

221.5
393.7
615.2
885.9
1205.8
1574.9
1993.2
2460.8

332.2
590.6
922.8
1328.8
1808.7
2362.4
2989.9
3691.2

QT = Qf a + Qpa , t
Static
Dynamic

cm

kv
t/cm

332.0
541.1
799.4
1106.9
1463.7
1869.6
2324.8
2829.2

1.79
2.12
2.47
2.83
3.20
3.56
3.93
4.30

185.83
255.03
323.22
390.74
457.80
524.52
591.01
657.32

462.3
764.0
1139.6
1589.0
2112.2
2709.2
3380.1
4124.8

Symbols: Qfa Admissible frictional bearing capacity; Qpa Admissible point bearing capacity; QT, Total load
bearing capacity; Settlement; kv Modulus of vertical reaction. These resistances shall be compared against
the service load.
3.9 Bearing capacity for cut-off wall.
Both, the point bearing and the frictional capacity of
the cut-off wall (known as Milan wall in Mexico) were
calculated (resting at a depth of 28.0 m), by means of
expressions 1 and 2, respectively. The results show

an admissible frictional bearing capacity of 173.1 and


203.6 t/m whereas the admissible bearing capacity
varies between 100.0 and 200.0 t/m. In addition,
calculation was made of the negative skin friction
(FN) to be developed along the cut-off walls at

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

88

Geotechnical design of the foundation for an office building located at the transition zone

depths from 0.0 to 15.5 m through the use of the


following equation:

FN 0.3

with

o dz

(6)

where: , perimeter of the cut-off wall, m; and o dz,


integral of the effective stress at the length where the
negative skin friction is developed, 183.75 t/m.
Substituting the corresponding values, the
resulting value of the negative skin friction becomes
FN= 55.0 t/m.
3.10 Net admissible load bearing capacity for cut-off
wall.
With the previous results, the net admissible load
bearing capacity of the cut-off wall is equal to 217.9
and 348.6 t/m, for static and dynamic conditions,
respectively.

1 sin
1 sin

where:
Pd,
horizontal
pressure
envelope
corresponding to the active condition, t/m2; i, unit
weight of each stratum, t/m3; hi, thickness of each
stratum, m; ci, undrained cohesion for each stratum,
t/m2; Ni, factor depending on the angle of internal
friction corresponding to each stratum; ui; pore water
pressure, t/m2; qi, surcharge at surface, 2.0 t/m2; D,
depth of excavation, 24.0 m; and , angle of internal
friction, in degrees.
4.2 Passive horizontal pressures.
The passive horizontal pressures acting at the
internal part of the wall reacting against the soil were
evaluated as follows (Ref. 4):

hi i h i N i 2 ci

3.11 Settlements of cut-off wall.


The settlements were calculated with equation (4);
their magnitude became equal to 2.9 cm.

(8)

N i

(9)

where the symbols have been previously defined.


4.3 Long-term horizontal pressures.

3.12 Modulus of vertical reaction for the cut-off wall.


The modulus of vertical reaction was determined
from expression (5) and its magnitude amounts to
kvMiln= 93.51 t/cm.

The diagram of horizontal pressures to be withstood


by the cut-off wall was calculated by means of the
following expression (Ref. 4):
Ph K o ( i z q i u) u

(10)

where: Ko, at-rest pressure coefficient; the other


symbols have been already defined.

4 SHORT- AND LONG-TERM HORIZONTAL


PRESSURES

4.4 Seismic force.


4.1 Short-term horizontal pressures.
For purposes of design of the peripheral retaining
wall a maximum depth of excavation of 24.0 m was
considered, with cisterns included. Calculation was
made of the horizontal stresses associated to the
mass of soil; the envelope of the active horizontal
pressure (per linear meter of width) to be developed
against the wall was determined using the criterion of
Terzaghi-Peck and applying undrained shear
strength parameters corresponding to the most
unfavorable condition (Ref. 4):

Pd

2 ci
1.25 i h i u i
q
(

) dh i u i
D
N i
N i
N i

The seismic force, Fsis,to be developed in the soil


mass against the peripheral walls was evaluated
pursuant to the Complementary Technical Standards
applicable to seismic design; its magnitude was
calculated as follows (Ref. 2):
Fsis

W 4 ao
3

(11)

where: W, weight of the active wedge (considering


the surcharge of 2.0 t/m2), t; and ao, ordinate of the
design spectrum for time T=0, equal to 0.08.
After substituting the corresponding values the
diagrams of short-term and long-term horizontal
pressures were plotted as shown in Fig. 2.

(7)

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

ARENAS F. et al.

89

Figure 2. Diagram of short-term and long-term horizontal pressures.

E p a M pt

5 DESIGN OF THE EXCAVATION AND OF


ANCHOR SYSTEM

FS

To excavate the enclosure accommodating the


basement levels it will be necessary to stabilize the
wall by means of a retaining system constituted by a
0.40-m thick reinforced concrete cut-off wall founded
at a depth of 28.0 m complemented with five rows of
anchors.

where: Ep, passive earth pressure, t/m; a, lever arm


of the resultant of passive earth pressure at the
bottom level of anchors, m; Mpt, plastic moment of
sheet pile, 35.0 t-m/m; Ea, active earth pressure, t/m;
d, lever arm of the resultant of the active pressure at
the bottom level of anchors, m.
Substituting the corresponding values a safety
factor of about 2.0 is obtained.
Based on the diagrams of short-term horizontal
pressures the design of the anchor system was
accomplished to complement the peripheral retaining
wall during the excavation works and construction of
the foundations; it involved five anchor levels. Fig. 3
shows a proposal for the layout and distribution of
anchors.

5.1 Review for bottom failure.


The safety factor against plastic flow (creep) at the
bottom of the excavation (bottom failure) was
determined using the criterion of bearing capacity
assuming a prism of soil gravitating at the elevation
of the excavation bottom
having a width Be
representing a hypothetical footing:

FS

0.5 Be N
Pd q

(12)

where: , unit weight of the soil below excavation level,


1.65 t/m3; Pd, total vertical pressure at elevation of the
excavation bottom, 37.4 t/m2; q, superficial surcharge,
2.0 t/m2; Nbearing capacity factor, 44; Be, width of
hypothetical footing, 5.65 m.
Substituting the corresponding values a safety value
exceeding 2.0 is obtained.
5.2 Review for failure of embedment of cut-off wall.
Determination was made of the safety factor against
embedment of the cut-off wall upon completion of the
bottom level of struts and the maximum depth of
excavation has been reached; the expression used
reads as follows:

Ea d

(13)

5.3 Capacity of post-tensioned anchors.


The capacity of the anchors was determined from the
following expression:

Qf

Pi w le tg
FS

(14)

where: Qf, frictional capacity of anchor, t; Pi, grouting


pressure, 6.0 kg/cm2; w, perimeter of the cross
sectional area (15 cm in diameter), 0.47 m; le, length
of the anchor bulb, m; , angle of internal friction of
soil where the anchor bulb will be embedded, 36
degrees; and FS, safety factor equal to 1.5.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

90

Geotechnical design of the foundation for an office building located at the transition zone

5.4 Characteristics of anchors.

6 INSTRUMENTATION

Their diameter will be of 15 cm and their bore holes


will be drilled with pneumatic equipment, having an
inclination of 45 degrees below the horizontal; they
are constituted by 0.6-diameter strand cable. The
annular void between the borehole walls and the
strand is filled with a water-cement grout with
proportion of 2:1 by weight with minimum strength of
150 kg/cm2, injected at a pressure of 6.0 kg/cm2. To
quickly reach the strength of the grout a setting
accelerator will be used and a volume stabilizer.
Within a term of three days, when the grout has
already reached the design strength, post-tensioning
(Fp) specified in tables 2 and 3 was applied. The
reaction system of the anchors was resolved with
concrete blocks supporting the reaction and wedging
plates.
Table 2. Anchor system for panels of 2.5m, 4.0m and
5.5m.
Row

6.1 Objective.
The instrumentation program necessary to monitor
the movements of adjacent buildings and of excavated
area itself was defined as follows:
6.2 Bench mark.
For purposes of determining with accuracy the
ground movements to be experienced as a result of
the building construction, it is convenient to install a
control point outside the influence area of the works
to be carried out so that when reference is made to
these points the movements generated when building
of the foundation can be calculated as differentials.
Such bench mark shall be located at a distance of no
less than 300 m from the job site.
6.3 Measurement program.

Passive length,

Bulb length,

Total length,

Fp,

No. of

strands

22.0

5.0

27.0

43.0

16.0

8.5

24.5

110.0

12.5

9.5

22.0

131.0

9.0

6.5

15.5

70.0

6.4 Superficial references and wall marks.

6.5

5.0

11.5

54.0

They are constituted by points fixed at ground


surface and as reference points painted in
neighboring structures; the former are installed by
defining collimation lines parallel to the excavation
axes that are monitored with a builders transit so as
to detect the horizontal displacements that have
occurred, whereas with an optical level and stadia
rods the vertical displacements are measured.

Table 3. Anchor systems for panels of 3.4m, 6.0m and


7.0m.
Row

Passive length,

Bulb length,

Total length,

Fp,

No. of

strands

22.0

5.0

27.0

51.0

16.0

9.5

25.5

132.0

12.5

11.5

24.0

157.0

10

9.0

6.5

15.5

84.5

6.5

5.0

11.5

64.5

Reading of the reference point shall be made once a


week with results presented graphically to facilitate
their interpretation.

6.5 Reference points in cracks.


Gypsum marks were plastered to monitor the
enlargement of cracks and fissures at those sites
where they were detected by the structural inspection
of neighboring buildings. On the other hand, a survey
was made of cracks at sidewalks and street
pavement to monitor their behavior.

N.P.T +4.50

Planta Comercial

N.P.T +0.00

Estacionamiento S - 1

N.P.T -4.71

Estacionamiento S - 1

Estacionamiento S - 2

Estacionamiento S - 2

N.P.T -7.85

Estacionamiento S - 3

Estacionamiento S - 3

N.P.T -11.00

Estacionamiento S - 4

Estacionamiento S - 4

N.P.T -14.14

Estacionamiento S - 5

Estacionamiento S - 5

N.P.T -17.28

Estacionamiento S - 6

Estacionamiento S - 6

N.P.T -20.42

Estacionamiento S - 7

Estacionamiento S - 7

N.P.T -23.57

Figure 3. Schematic cross section and a front view of the layout proposed for the anchor system.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

3rd.

International Conference on Deep Foundations


Deep Foundations and Soil Improvement in Soft Soils

Session 2:
Excavations

Technical Committee

TC-214

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

A historic capitol and a deep excavation


Un capitolio histrico y una excavacin profunda
Nasser MASSOUDI1 & Richard SLIWOSKI 2
1Bechtel
2Virginia

Power Corporation, Frederick, Maryland, USA


Department of General Services, Richmond, Virginia, USA

ABSTRACT: The Virginia State Capitol was designed by Thomas Jefferson and constructed in 1785. The Capitol is of
great historic and architectural significance to not only the Commonwealth of Virginia but to the nation. It is the oldest
operating Capitol in the U.S. and required a total restoration/renovation as well as an expansion to continue to remain
functional. Site conditions dictated that the expansion (a new visitor center) be placed underground on the south side of
the Capitol building, requiring a 40-foot deep excavation approximately 5 feet from the Building. Because of its historic
significance, every measure had to be taken to ensure the safety and function of the Capitol during construction. The
construction techniques that were used to make the deep excavation and protect the historic Capitol included a tied back
concrete slurry wall, jet grouting, compensation grouting, and instrumentation monitoring. This paper describes the
history of the Capitol building, the techniques that were employed for construction of the deep excavation, results of the
latest movement monitoring, and factors contributing to meeting the restrictive movement goals.

1 HISTORICAL BACKGROUND
The Virginia Capitol in Richmond houses the oldest
legislative body in the United States. It has played a
significant role in Virginias history with the
contributions of its many historical figures as part if
its tapestry. It is designated a National Historic
Landmark and is listed on the National Register of
Historic Places.
Virginias first Capital was in Jamestown and dates
back to 1619. The State Capitol was relocated to
Middle Plantation (Williamsburg) in 1699. It served
until the American Revolutionary War. It was
Governor Thomas Jefferson who urged that the
Capitol be relocated to Richmond. In 1779, the
Virginia legislature voted to move the Capitol from
Williamsburg to Richmond. Plans soon began for a
new building to serve a new state, the
Commonwealth of Virginia.
With the establishment of Richmond as the new
capital, six squares of land were selected for the
placement of permanent public buildings on
Shockhoe Hill, a major hilltop overlooking the falls of
the James River in Richmond.
Thomas Jefferson designed the Virginia State
Capitol, which is the middle structure of the present
Capitol building, while serving as minister in France.
Working with French draftsman Charles-Louis
Clrisseau, Jefferson designed the building to

represent a dramatic departure from British influence.


He modeled the building after the Maison Carre at
Nmes in southern France, an ancient Roman
temple. Jefferson set a precedent by using a temple
form and the building is nationally significant as the
first Classical Revival state capitol building
constructed in America. He originally intended to
erect three buildings, one for each of the three
branches of government. However, his goal was
beyond the Commonwealth's finances.
On August 18, 1785 the cornerstone was laid for
the Capitol building in Richmond. However,
construction of the Capitol had begun without the
plans. Jefferson sent the Clrisseau drawings and
the plaster model created by Fouquet; they reached
Richmond separately in 1786. The original model is
still on display at the Capitol building and is shown in
Figure 1.
Once the building plans arrived, it was discovered
that foundations had been laid out different than that
in Jeffersons plans. During the long period of
construction, 1785 to 1798, Jeffersons design was
extensively altered by Samuel Dobie and other
Richmond builders, including the foundations.
Thomas Jefferson had expressed unhappiness with
the changes in his design.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

94

A historic capitol and a deep excavation

Figure 1. Original Capitol Building Model.

The late 1700s was just the beginning for the new
state house as the location of many extraordinary
moments in history. In December 1791, the Bill of
Rights was approved making Virginia the 11th state
to ratify the amendments; and in 1796 Jean-Antoine
Houdons statue of George Washington was placed
in the rotunda becoming one of Virginias most
treasured artifacts.
By 1857, the building was
suffering from deferred maintenance and the effects
of heavy use. Unfortunately, the cost of renovation
was deemed too high and repairs were substituted.
In 1858, a proposal to enlarge the Capitol building
was submitted by Albert Lybrock. However, in 1861,
the Virginia Convention voted to secede from the
union. The American Civil War, in which Virginia
played an important role, interrupted the project and
Lybrock's proposals were never executed. In 1862, in
the House of Delegates Chamber, Robert E. Lee was
appointed as commander of the Army of Northern
Virginia. The Capitol building would serve as the
Capital of the Confederate States of America from
1861 to 1865. In April 1865, departing Confederate
troops were ordered to burn the citys warehouses
and factories. The fire spread out of control and the
Capitol building was one of few buildings that were
spared, as shown in Figure 2.

African-American building contractor, offered a


resolution to address the appearance and condition
of the Capitol. During this period, there was a
dispute over the leadership of the City of Richmond
government. This led to a hearing on April 27, 1870
that was held in the large courtroom on the second
floor of the Capitol building. Several hundred people
crowded the room and balcony. The balcony was
overloaded and the extra weight caused it to collapse
and fall about 40 feet to the courtroom floor below.
This soon came to be known as the Capitol
Disaster which caused 62 deaths and injured 251
people. Despite calls for the building's demolition,
the damage from the tragedy in 1870 was repaired.
By the turn of the century, the Capitol was in a
poor state of disrepair. Therefore, the General
Assembly appropriated $250,000 for the Capitol
renovation. After 16 months of work, the remodeling
was completed, consisting of the addition of the
wings to the east side of the Capitol building (House
Chamber) and west side (Senate Chamber),
including the addition of 24 granite steps on the
South Portico of the building; the latter was in
keeping with Jeffersons original plan. This significant
renovation from 1904 to 1906 was designed by noted
Norfolk Architect John Kevan Peebles. Additional
interior work ensued later, in 1937 and 1962.
The latest renovation of the Capitol started in 2003
and is scheduled for completion in late 2006. This
almost $100 million project includes the renovation
and restoration of both the exterior and interior of the
Capitol, and the construction of a 27,000 square foot
extension below ground, connecting to the Capitol.
Upon completion, the Capitol will have new
mechanical and electrical systems; a new roof,
elevators, stairwells, and legislative meeting space;
the building will be fully accessible and ADA
compliant; the legislative chambers and Rotunda will
have been restored to their original color and dcor
of the time period; and there will be a new visitor
entrance at 10th and Bank Streets with a new visitor
center in the underground extension providing
educational and museum exhibits. The exhibits are
scheduled to be opened in 2007, at the time of
Virginias 400th anniversary.
2 DEEP EXCAVATION

Figure 2. Capitol Building During the Civil War.

During the Reconstruction period, Virginia was


under military rule. By 1870, the Capitol was overdue
for repair and Delegate Ballard Edwards, a respected

From a geotechnical perspective, the most significant


aspect of the recent construction is related to the
installation of a 40-foot deep excavation as close as
5-foot from the Capitol building to house the new
visitor center. This excavation becomes even more
significant considering the requirements that it had to
meet, including protecting the historic Capitol, which
rests on soft deteriorated footings that are over 200
years old, and by allowing only very limited building
movements of 0.25 inch. The condition of some of
the footings is shown in Figure 3.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MASSOUDI N. et al.

Capitol

95

Building

Capitol Foundations

Slurry Wall

E x c a v a t i o n

A r e

Figure 3. Original Capitol Footings.

3 GEOLOGY
Virginia Capitol is located on a hilltop. The site
geology is characterized by sequences of marine and
sedimentary deposits. The site soils consist of manmade fill, natural soils of various geologic ages, and
bedrock. Typical subsurface stratigraphy at the site
consists of approximately 2 feet of medium dense
sandy fill, 20 feet of loose to very dense clayey sand
with cobbles and boulders (average Standard
Penetration Test N-value of 40 blows/foot), 70 feet of
firm to hard clay and silt (average N-value of 15
blows/foot and average undrained strength of 2,000
pounds per square foot), 50 feet of very dense sand
and gravel (SPT values typically greater than 100
blows/foot), and granite bedrock at a depth of about
140 feet below the ground surface with average
unconfined compressive strength of 15,000 pounds
per square inch. Perched groundwater level is about
20 feet below the surface. Only soils of upper 80 feet
are of geotechnical significance for the project,
consisting of clayey sand, and clay and silt, as these
soils have the greatest influence upon the stability of
the construction.
4 EXCAVATION SUPPORT CONSTRUCTION
The main component of the excavation support
system was a reinforced concrete slurry wall, with a
total length of about 150 feet. It was 60 feet deep
from the existing ground, with a thickness of 3.3 feet.
It served both as a temporary excavation support and
as the final structural wall for the new visitor center.
The geometry of the wall in plan was unusual,
dictated by architectural requirements. The slurry wall
was tied to 4 transverse sections; these sections
were designed as cantilevers. The slurry wall and its
relation to the Capitol foundations are shown in
Figure 4.

Figure 4. Slurry Wall in Plan.

Given very limiting movement requirements for the


project, it was found necessary to improve the soils
between the Capitol foundations and the slurry wall
using jet grouting. The grouting was only in the upper
20 feet of the ground. More than 100 jet grout
columns were installed. Each column was nominal 3
feet in diameter and was keyed into the underlying
clay.
Similarly, compensation grouting was deemed
necessary as a precaution against excessive
movements. Grouting was performed in the upper 20
feet of the site soils immediately under the Capitol
foundations. The Capitol building was occupied
during this work; therefore, all grouting had to be
performed from outside the building, including
grouting underneath the interior Capitol foundations.
With the slurry wall and other grouting components
in place, excavation in front of the slurry wall
commenced, supported by 5 rows of temporary
tiebacks one of which was drilled directly into the jet
grout columns and the remainder through the slurry
wall. The tiebacks consisted of three to six 7-wire
strands. All tiebacks were post-grouted, pre-stressed,
and tested. Tiebacks in the jet grout columns were
stressed to 70 kips each, transmitting the
prestressing load through steel walers to the jetgrouted mass. Tiebacks in the slurry wall were
stressed to loads varying from 60 to 160 kips. As
noted earlier, the tiebacks were only temporary; once
the structure of the visitor center was completed all
tiebacks were de-tensioned.
5 GEOTECHNICAL INSTRUMENTATION
The extraordinarily limiting movement goal for the
project of 0.25 inch required the availability of
performance information on a continual basis to
enable rapid response to developing trends. It was
also a critical part of the verification and quality
assurance. Hence, a real-time system of
geotechnical instruments was used for monitoring the
building and the construction. They consisted of three
total station theodolites with about 80 optical prisms,

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

96

A historic capitol and a deep excavation

two in-place inclinometers (and one manual as backup), temperature sensors, and data acquisition
hardware, software, computers, and Internet
connection.
The system proved invaluable in monitoring the
performance of the work. At one time during the
installation of the jet grout columns in November
2004, several Capitol footings began to settle
unexpectedly. The availability of continuous and realtime results enabled a quick response in averting
excessive foundation movements with undesirable
consequences. These movements were later
recovered. Equally, the real-time data was
instrumental in controlling the grouting operation
during the compensation grouting and tieback
installation. Typical vertical movement monitoring
results are shown in Figure 5.
0.25

Target OWT0111

Movement (inch)

0.2
0.15
0.1

Jet grouting period

0.05
0
-0.05
-0.1

excavation support, ground improvement using jet


grouting,
foundation
improvement
using
compensation grouting, and a very restrictive
movement tolerance of 0.25-inch for the Capitol.
With the deep excavation complete and construction
of the new underground visitor center underway, the
final movements are well within the goal of 0.25 inch.
The maximum movement recorded during the entire
construction period was 0.17 inch of settlement which
occurred during the jet grouting operation. This
settlement, however, has since been recovered, with
final movements largely in the 0.1-inch range. The
established movement tolerances for the project
were extraordinarily limiting compared to typical
excavations, yet fitting and consistent with protecting
deteriorating foundations and historic architecture.
The achieved movement results are also
extraordinary for the given excavation depth, perhaps
including an element of good fortune. The success in
controlling movements is credited to detailed design,
close communication among the project team,
proven construction practices, and a real-time
monitoring system that permitted the evaluation of
construction performance at close intervals and
implementation of corrective measures within very
short time periods.

-0.15
-0.2
-0.25
Sept Dec
Mar
Jun
Sept Dec
Mar
Sep-04 Nov-04 Feb-05 May-05 Aug-05 Nov-05 Feb-06
04
04
05
05
05
05
06

Date

Figure 5. Typical Movement Monitoring Results

Results of movement monitoring for over 18


months of construction are shown in Figure 5. With
structural components for the new visitor center
completely in place, the Capitol building movements
to date are well within the goal of 0.25-inch. The
maximum recorded movement was 0.17 inch of
settlement as observed during the jet grouting period.
However, this settlement was later recovered, with
final movements largely in the 0.1-inch range.
6 CONCLUSIONS
The Virginia State Capitol was designed by Thomas
Jefferson and is the oldest operating Capitol in the
United States. With great historic and architectural
significance, it is recognized as the model for public
buildings throughout the United States. Having
survived the American Civil War and many years of
deterioration, it is undergoing a much needed
restoration as well as an expansion to continue to
remain functional. The construction of a 40-foot deep
excavation exposed the Capitol to risks of excessive
movements beyond that considered tolerable.
Therefore, redundant construction measures for risk
mitigation were implemented for meeting the
challenge. These included a concrete slurry wall for
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

The support of a 25 m deep excavation in difficult ground conditions using


Single Bore Multiple Anchor technology
Soporte de una excavacin de 25 m de profundidad en condiciones de terreno difcil usando
tecnologa de anclaje mltiple con barreno nico
Devon MOTHERSILLE1, and Bora OKUMUSOGLU 2
1Single

Bore Multiple Anchor Ltd, London, UK


2Kasktas AS, Moscow, Russia

ABSTRACT: The paper describes the design, construction and testing of some 3600 temporary single bore multiple
anchors (SBMAs) used to support the deep basement which forms part of the foundation for the Kuntsevo Plaza; a
mixed-use development in the Kuntsevo district of south-western Moscow. SBMAs were used to provide support for 40m
(131ft) deep diaphragm walls, constructed to retain a 25m (82ft) deep excavation in the challenging Moscow mixed soils,
comprising combinations of low strength clays, sands and silts. Previous attempts to sustain the required loads of up
600kN (135kips) in the anchors had failed due to unacceptable creep. However, an understanding of the concept of
progressive debonding, and the use of this knowledge in the design of efficient fixed anchor lengths, in the SBMAs,
proved highly effective. In addition, the introduction of fixed anchor enhancement techniques such as post-grouting
resulted in the achievement of anchor capacities more than double those previously achieved in the prevailing ground
conditions and proved effective in limiting wall displacements to a maximum of 7.5mm. This project was also the genesis
of intuitive and innovative cloud-based software, specifically developed for tablets, to analyze and manage the vast
volumes of data produced from the stressing and testing of the ground anchors.

1 INTRODUCTION
Ever since the first development plan in the 16th
Century, the city of Moscow has drafted and
implemented several development strategies over
the centuries which have contributed to a city
steeped in history and culture. More recently, in a bid
to attract new investors to the city, an urban
development plan was drafted that will take Moscow
up to the year 2020. Kuntsevo Plaza, completed in
late 2014, forms part of this strategy. The Plaza,
located in a south-western district of Moscow (Figure
1), is described as a vibrant mix-use lifestyle centre
rooted in urbanity and comprising five levels of
amenities above the surface including offices,
residential, retail, restaurants and leisure facilities.
An essential feature of the new construction and
the main focus of this paper is the foundation
structure which supports a 25m deep excavation.
The excavation accommodates the vast basement of
the complex where features include extensive car
parks, utilities and a substantial cinema complex. In
order to facilitate construction of the basement a
diaphragm wall of 45m total depth, 0.8m thick and
approximately 600m
overall perimeter was
constructed.

Figure 1. Location plan showing Kuntsevo Plaza in a


south-western district of Moscow.

2 GROUND CONDITIONS
Two site investigations were carried out to establish
the nature, and more importantly, the engineering
properties of the ground. The extent of the site
together with trial anchor, borehole and CPT
locations, which are referred to later in this paper, are
shown in Figure 2. Boreholes were generally driven
to depths of up to 50m and some 605 samples were
extracted and tested, and these included samples
used to establish the aggressivity of the ground and
the ground water.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

98

The support of a 25m deep excavation in difficult ground conditions using Single Bore Multiple Anchor
technology

ultimate bond stresses at the ground/grout interface


is particularly useful from a fixed anchor design
perspective but it is emphasised that the magnitude
of this parameter is also a function of the contractors
construction methodology.
3 THE APPLICABILITY OF THE SBMA CONCEPT

Figure 2. Simplified site plan showing trial anchor,


borehole and CPT test areas.

The final design for the anchored structure utilized


ground anchors installed across six levels. In the
vicinity of the fixed anchors, associated with the first
four levels, mixed soils comprising clay interbedded
with sand, glacial sandy silt with lenses of gravel and
fluvio-glacial deposits of water-saturated fine to
medium sands were encountered. In the lower two
levels, the fixed anchor zones were founded in Lower
Cretaceous deposits of up to 20m thick water bearing
sands with clay pockets. These sandy strata were
underlain by impermeable Upper Jurassic clay
deposits, in which the diaphragm wall is socketed.
The natural ground elevation varied between 170.00
to 174.00m above sea level and the groundwater
elevations varied between 164.72 and 167.10m, with
two additional artesian aquifers located in the fluvioglacial and Lower Cretaceous sands.
The geotechnical parameters established from
laboratory tests included quantification of effective
angle of shearing resistance, drained Youngs
Modulus, effective cohesion, and unit weight of soil.
These were relevant to varying degrees for the
design and the computer modelling of the anchored
structure, and are listed later in this paper in Table 2.
It was established that the specialist anchor
contractor, Kasktas A S, had undertaken a series of
field trials on tremie grouted conventional anchors,
which comprised anchors with a single 8m bond
length, installed vertically into the two distinct
founding layers. Borehole diameters of 150mm and
178mm were used and drilled to depths of between
18m and 27m. During these trials, the ultimate load
was defined as the maximum load attained by the
anchor before continuous upward displacement of
the fixed anchor was observed. Test data confirmed
that this generally occurred at 450kN. Back analysis
of the failure loads generated average ultimate bond
stresses at the ground/grout interface of 175 kPa and
195 kPa in the upper clay and in the lower sand
layers, respectively. The establishment of in situ

Recognizing that conventional straight shafted


grouted anchors failed to achieve the desired working
loads in the prevailing ground conditions, alternative
technology had to be implemented. With this
background SBMA Ltd were approached by the
specialist anchor contractor to design ground
anchors that could fulfil the design requirements.
It has been acknowledged by numerous
researchers for over 60 years that when tensile load
is applied to a steel tendon in grout, whether founded
in rock or soils, the load distribution within the fixed
anchor length is non-uniform.
Ostermayer (1974 and 1977) and subsequent
work by Barley (1995) into the performance of
anchors founded in clays, sands and gravels
highlighted the non-uniform distribution of bond
stress and the progression of load concentration
along the length of long fixed anchors. As a
consequence the British Standard code of practice
for Ground Anchorages (BS8081:1989) recommends
a limit of 10m for the fixed anchor length.
Barley (1995) established a method of evaluating
the efficiency of an anchor in mobilizing ground
strength; short fixed lengths being highly efficient in
mobilizing ground strength and long fixed length
being grossly inefficient (i.e. a 10m fixed anchor
utilizes only 45% of the average bond stress
exhibited in a 2.5m fixed anchor).
Furthermore, and most importantly, the utilization
of a multiple of short and efficient unit lengths within
a single borehole has allowed the working capacity of
soil anchors to be more than doubled. Working loads
in the range 600 kN to 3800 kN have been safely
achieved in ground conditions ranging from soils to
highly weathered rocks in various parts of the world
and this methodology was implemented in the design
of the Kuntsevo ground anchors.
4 SBMA - DESIGN, FABRICATION AND
CONSTRUCTION
The SBMAs were designed in accordance with
procedures documented in Ostermeyer and Barley
(2003) which incorporates an efficiency factor to
account of the non-linear distribution of bond stress,
due to progressive debonding, that exists in the
grouted tendon under increasing tensile load. A
simple mathematical expression relating the ultimate
bond stress, ult, to the length of the fixed anchor
was applied;

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MOTHERSILLE D. et al.

ult

L x feff

where feff is an efficiency factor, which itself is a


function of L in the form; 1.6L-0.57 and L is fixed
anchor length (m).
The tendons, incorporating 7- wire low relaxation
strand (Characteristic tensile strength = 261 kN) and
greased HDPE sheathing for decoupling the free
length, were fabricated on site ensuring that
dimensions and preparation of materials were carried
out in accordance with the relevant standards for
temporary anchors. The borehole was formed
through a steel-reinforced concrete stressing block
(2.0m x 2.0m x 0.5m deep) for investigation tests,
and for production anchors through reservation pipes
placed within the reinforcement of the diaphragm
wall. Boreholes were drilled with 145mm diameter
auger drill with end-of-casing air flush using a
diesel/hydraulic rig (Casagrande Type C8). In order
to preserve the integrity of the borehole a steel
casing (180mm o/d and 160mm i/d) was advanced to
the full depth of hole. On completion of the borehole
the complete tendon, comprising the three unit
anchors [top (A), middle (B), bottom (C)], the primary
grouting pipe and the post grouting pipe were
installed through the casing (Figure 3).

99

applications) the document providing guidance on


ground anchor practice in Russia.
In relation to investigation tests, EN1537:1999
(recently
revised
to
EN1537:2013)
states
Investigation tests may be required to establish for
the designer, in advance of the installation of the
working ground anchors, the ultimate load resistance
in relation to the ground conditions and materials
used, to prove the competence of the contractor
and/or to prove a new type of ground anchor by
inducing a failure at the ground/grout interface.
The investigation tests were carried out in phases
across three separate locations on the site (Figure
2). The three trial areas were selected based on
availability of suitable areas and logistics imposed by
ongoing construction activities. Mothersille et al
(2012) describe the first phase of the investigation
tests which was carried out in trial area 2. The main
objective of each testing phase was to assess the
performance of the SBMAs using different grouting
techniques in different locations on the site. In
addition the tests provided the contractor an
opportunity to become competent in the fabrication
and execution of SBMAs which had never been used
before in Russia. A series of fifteen test holes were
proposed to accommodate different types of test, the
details of which are summarized in Table 1 and
reference can also be made to Figure 2 which shows
the location of the test areas.
Table 1. Summary of trial anchor details.

Figure 3. Schematic showing SBMA with three unit bond


lengths (the primary grout pipe is omitted for clarity).

5 INVESTIGATION TESTS AND TRIAL ANCHOR


PROGRAMME
Bearing in mind the complexity of the ground, the
importance of the project, the high consequences of
failure and the fact that a new anchoring technology
was introduced to the specialist contractor, it was
considered prudent to carry out an extensive
programme of investigation tests. The specification
for the works referred to both EN1537:1999
(Execution of special geotechnical work ground
anchors), BS8081:1989 (British Standard Code of
practice for Ground Anchorages) and VSN 506-88
(Design and installation of ground anchor

Trial
anchor
No

Test
Area/
Phase

1/I

Description

Post-grouting test. 25m borehole inclined


at 15
2
1/I
29.5m SBMA, inclined at 15 using KECG
3
1/I
25m SBMA, inclined at 15 using TAM
4
1/I
15m SBMA, inclined at 15 using TAM
5
1/I
25m SBMA, inclined at 15 using CTG
6
2/II
Post-grouting test. 25m borehole inclined
at 15
7
2/II
29.5m SBMA, inclined at 15 using KECG
8
2/II
25m long SBMA, inclined at 15 using
TAM
9
2/II
15m long SBMA, inclined at 15 using
TAM
10
2/II
25m long SBMA, inclined at 15 using
CTG
11
3/III
Post-grouting test. 25m borehole inclined
at 15
12
3/III
29.5m SBMA, inclined at 15 using KECG
13
3/III
25m SBMA, inclined at 15 using TAM
14
3/III
15m SBMA, inclined at 15 using TAM
15
3/III
25m SBMA, inclined at 15 using CTG
*KECG denotes Kasktas end-of-casing grouting
**TAM denotes tube-a-manchette
***CTG denotes conventional tremie grouting

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

100

The support of a 25m deep excavation in difficult ground conditions using Single Bore Multiple Anchor
technology

5.1 Anchor grouting strategies


Post-grouting of anchors is a well-established
technique and known to significantly enhance the
capacity of anchors given the appropriate ground
conditions. Ostemeyer and Barley (2003) note that an
enhancement factor of two can be reasonably
achieved on post-grouted SBMAs installed in clay
soils. However, the challenge confronting many
contractors
is
deciding
what
post-grouting
parameters should be applied for a particular type of
ground. Often they have to rely on experience or trial
and error techniques in order to generate the
enhancement in capacity desired.
To gain an understanding of the ground response
to post-grouting and produce a rational basis for the
execution of post-grouting in the production anchor
works, a series of three post grouting trials were
included in the investigation test programme. These
were designed with the objective of establishing a
range of break-out pressures for the tube-a
manchette (TAM) valves, refusal pressures, volume
of grout take and flow rates. Such parameters are a
necessary prerequisite to establishing an effective
post-grouting regime in the production anchors.
The post grouting trial involved the installation of
25m long TAMs, eccentrically placed in the bore hole
to simulate the position they would adopt were an
anchor tendon placed in with them. The TAM
installation comprised approximately 10m of
perforated tubing, with ports spaced at 300mm
centres (representing the unit anchor locations), and
15m of riser pipes, connecting the perforated tubes
to the surface. The installation was oriented subvertically at an angle of 15 to the vertical axis.
The trial work established a methodology which
involved the use of 50mm diameter TAMs and an
injection of a target volume of grout of 20 litres per
port (spaced at 300mm c/c) with water/cement (w/c)
ratio of 0.6 and an associated injection pressure of
10 bar. This was the starting point and crucial
grouting parameters were monitored using a
computerized Atlas Copco Logac G5 recorder
system which provided data on pressure, flow rate
and volume. The data derived from these tests
assisted the contractor with the planning of the
grouting strategies for the production anchor works.
When sandy soils or substantially non-cohesive
strata were encountered end-of-casing grouting
techniques were employed. This method typically
involves driving casing through to the end of the
borehole and then applying a controlled pressured
grout (typically 10-20 bars) through the casing as it is
incrementally withdrawn and has proven effective in
creating enhancement of bond capacity by a factor of
2 (Ostermeyer and Barley, 2003).
Conventional tremie grouting was also employed
to compare the behaviour of the test anchor
incorporating other grouting techniques. The primary

grout retained a w/c of 0.45 for both end-of-casing


grouting and tremie grouting.
5.2 Stressing and testing
The jack arrangement for a three-unit SBMA includes
three hydraulic rams that are synchronized by
coupling to the same hydraulic powerpack, so that
the same load is applied simultaneously to each unit
anchor. The stressing and testing arrangement is
shown in Figure 4. The ram extensions on the
stressing jack were recorded using a digital vernier
caliper for each stage of the cyclic loading and again
during creep testing in accordance with BS8081.
Measurements were corrected for reaction base-pad
movement measured by dial gauges mounted on an
independent reference beam.

Figure 4. Typical set for the stressing and testing of


production SBMAs at Kuntsevo showing three
hydraulically synchronised stressing jacks.

5.3 Observations and salient points derived from


investigation tests
5.3.1

Tests loads adopted for trial anchors

At Kuntsevo the geometrical configuration of the trial


anchors was identical to that proposed for the
production anchors but for practical reasons the
anchor had to be installed sub-vertically as opposed
sub-horizontally. Furthermore, although production
anchors are classified as temporary anchors (which
are typically associated with a proof load factor of
1.25 and a maximum proof load of 665 kN) it was
suggested at the outset, by Kasktas, that the anchors
should be tested to 900 kN. Knowing that the
maximum design working load was 532 kN this
would mean imposing an onerous proof load factor of
1.7 on the anchors. The decision by Kasktas to use a
higher proof load factor was partly due to the acute
concern over minimizing settlements induced behind
the wall and the effects of these on adjacent buried

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MOTHERSILLE D. et al.

services and this formed the basis of a conservative


approach in the design. The trial anchors were
therefore provided with an increased tendon capacity
to allow them to be tensioned to a maximum test load
of 1250kN without exceeding 80% tendon capacity.
In order to assess the performance of the trial
tests, calculations were made to derive the maximum
permissible working loads for the founding stratum
encountered. The maximum permissible working
loads were derived using a limiting magnitude of
creep displacement (2mm) in accordance with VSN
506-88 and Appendix M.10 of BS8081:1989.
5.3.2

Trial anchors

Trial anchor Nos. 2, 7 and 12 were designed in order


to derive data from the stratum that would
accommodate the bonded length for the lower rows
of production anchors. Bearing in mind the nature of
the ground encountered (described as sands) these
trial anchors were grouted using end-of-casing
grouting techniques.
The end-casing grouting methods employed were
not conventional in that the equipment was not
available to allow pressure to be applied through
300mm lifts as the casing is rotated and withdrawn.
Instead, Kasktas applied pressure through the casing
each 1.5m length was withdrawn. This created
significant pressure dissipation to the surrounding
ground. Furthermore, the grout records confirmed
relatively low pressures (average 2.5 bar for w/c =
0.45 grout) which suggests limited enhancement for
a given volume of grout injection. This was reflected
in the load/extension results which show relatively
large permanent displacements (circa 100mm) in the
upper units and lower permissible working loads.
Trial anchor Nos. 3, 8 and 13 comprised 25m long,
three unit SBMAs with an integrated TAM. The TAM
was included here to provide a comparison on
performance with other methods and to allow the
operatives an opportunity to be familiar with the
operation of a double inflatable packer system within
the TAM. Based on previous experience it was
judged that this system would provide a reliable
method of controlling post-grouting episodes in the
various founding layers.
Trial anchor Nos. 4, 9 and 14 were designed to
assess the performance in the shallower anchors
(generally the upper two rows). An overall
assessment of the tests was that they performed
satisfactorily and based on the results that were
derived from the tests an average maximum working
load of 703 kN was assessed from back-analysis.
The maximum required working load for these rows
of anchors is 412 kN which was adequately
exceeded.
Trial anchor Nos. 5, 10 and 15 were grouted via a
conventional tremie pipe with primary grout of w/c =
0.45, and no enhancement strategy employed. The
intention here was to provide a yard stick to compare

101

the performance of the other grouting techniques


employed. The stressing data derived for these trial
anchors appeared more erratic than those acquired
from
the
other
tests.
Larger
permanent
displacements were observed together with nonreproducible cyclical profiles in the load/extension
curves. The fact that these anchors were grouted
using conventional tremie pipe with no facility for
enhancing capacity was clearly reflected in the
relatively poor performance observed.
Analysis of the grouting records for these trial
anchors confirmed a positive effect of post-grouting.
This was particularly apparent in the behaviour of
anchor No.13 where the permanent displacement
was reduced (as illustrated in Figure 5) in the lower
units which received 30 litres of grout in the even
numbered ports and 20 litres in the odd numbered
ports. The target volumes were selected based on
previous experience but based on these results it
could be argued that the target volume should be
increased. It is noteworthy that Bruce et al (2004)
adopted grout volumes of 55 litres per port and
associated target pressures of 800 psi (55 bars) for
anchors installed in clays.
Back-analysis of the derived data suggests
maximum working loads ranging from 580-780 kN
(with the exception of the trial anchor No. 15 which
only had the conventional tremie grouting) exceeding
the maximum working load of 532 kN demanded by
the anchored structure design.
5.3.3 Summary and considerations for production
anchor works
The permissible working load of SBMAs even without
pressure grouting, were found to be greater than the
ultimate capacity (bond failure) of conventional
anchors. The use of SBMAs increased the effective
load capacity for the complex soil conditions, while
pressure grouting improved the load displacement
performance further.
From an overall assessment of the results it was
concluded that the use of the TAM was a necessary
general requirement for all SBMAs installed on the
site based on the following;
Contemporaneous geotechnical data indicated
a significant cohesive component within the soil
matrix. This was not obvious from the original soil
descriptions but was confirmed when the spoil heaps
were examined after drilling. It is commonly accepted
that post-grouting is particularly effective in cohesive
soils.
In the event that a production anchor exhibits
excessive creep during testing, the existence of the
TAM provides a means of revisiting the anchor to
introduce further episodes of grout prior to re-testing.
The load/extension data for the trial anchors,
grouted via tremie pipe, confirmed this method to be
inappropriate for the conditions encountered.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

102

The support of a 25m deep excavation in difficult ground conditions using Single Bore Multiple Anchor
technology

Reliance could not be placed on the


effectiveness of the end-of-casing grouting methods
implemented by the specialist anchor contractor.
Based on the observation from the investigation
tests, the following recommendations were made for
the production anchor works:
The SBMAs to comprise 3 x 3m long unit
anchors each with two strands which were locally
noded to enhance the bond resistance at the
grout/tendon interface
The use of TAMs was recommended for
SBMAs installed in the mixed soils encountered at
Kuntsevo Plaza.
The target grout volume should be increased to
50 litres per port with target pressure of 55 bars.
The location of the TAM should be carefully
assessed and efforts should be made to maintain the
TAM, by flushing with water, for subsequent regrouting if necessary
6 PRODUCTION ANCHOR SUPPORT SYSTEM
AND LOCATION OF ADJACENT SERVICES
Figure 6 shows a section through the ground support
system for the 25m deep basement excavation,
which comprised a diaphragm wall, socketed in an
impervious clay layer, and 6 levels of temporary
SBMAs with up to 600 kN working load.
The footprint for the development demanded that
the diaphragm wall be located within 4m of a
reinforced concrete sewage pipe. The top 5m of the
wall was designed to cantilever due to the close
proximity of adjacent services as shown in the Figure
6. The services are defined as follows; i, r/c sewer
pipe (2000mm diameter); ii and iii, steel gas mains
(420mm and 300mm diameter); iv and vi, r/c storm
water drains (400mm and 600mm diameter); v, steel
water main (600mm diameter) and vii, r/c storm
water culvert (3400mm x 3600mm).
The primary function of the support system was to
minimize the adverse effects of the deep excavation
on the buried services, many of which were already
vulnerable due to their age.
During the design phase of the deep excavation
the serviceability displacement limits for these
services, especially the sewerage line, which was
closest to the wall, influenced the inclination of the
ground anchors, which were installed at 1m centres.
In addition, the anchors were prestressed up to 600
kN to effectively restrict wall displacements, thereby
reducing settlements behind the wall. So sensitive
was the issue of settlements that the local authorities
insisted that the main contractor consider the
relocation or costly replacement of the sewage pipe if
assurances could not be given that settlement would
be within specified maximum settlement tolerance of
30mm.

Figure 5. Illustration of the positive effect of post-grouting


(TAM) in contrast to conventional tremie grouting (CTG).

Figure 6. Typical retained section with CPT profile.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MOTHERSILLE D. et al.

7 PRODUCTION ANCHOR PERFORMANCE


The three unit SBMAs were stressed and tested
using three hydraulically synchronized jacks. The
setup duplicated that used for the trial anchors
except for the fact that a more portable stressing
stool arrangement was adopted as shown in Figure
4. The specification required that 1 in 10 of all
production anchors be subjected to suitability tests
(equivalent to performance tests in PTI DC35.1-14)
and the remainder to acceptance tests (equivalent to
proof tests in PTI DC35.1-14). The first cycle was
carried out with nominal one minute reading at each
load level, the second and the third cycles were
carried out with 15 minute observation periods at
proof load level for monitoring the displacement-time.
At the proof load, creep displacements were
generally observed to be nominal and well within 5%
of elastic displacement of tendon as required by the
standard.
Service load monitoring of the anchors was carried
out by undertaking lift-off checks at the lock-off load
(approximately 110% working load) in accordance
with criteria stipulated in BS8081. On the rare
occasions that anchor did not satisfy acceptance
criterion, additional post-grouting was employed and
the anchor retested until satisfactory creep was
observed or a stabilizing trend was established.
Checks were also carried out on the apparent
tendon free length to ensure that the seat of load
transfer was within acceptable limits stipulated in
BS8081:1989. There were occasions when the
apparent tendon free length fell marginally short of
the 90% design free length criterion (the lower limit)
specified. The reasons for this were related to friction
developed in the free length and under these
circumstances the anchor was subjected to two extra
cycles of loading to gauge repeatability. Although this
methodology was justified and recommended in
BS8081, concerns were expressed about the
definition of what constituted a repeatable loadextension curve. BS8081 does not provide guidance
on this so reference was made to the Australian
standard (RTA QA Specification B114:2007, clause
9.6.2). In accordance with this standard repeatability
is investigated by checking that extensions in the
third cycle are within 5% of those in second cycle.
One of the major challenges was attempting to
execute the works in freezing climatic conditions
which affected all aspects of the ground anchor
execution. When temperatures dropped below zero
down to -15C, special heaters were deployed so that
operatives could record load-extension data with
some level of comfort. During construction heated
water tanks were deployed and insulation jackets
used on grout pipes. Below -15C the site operations
related to ground anchoring were abandoned. The
inclement weather accounted for a total of two weeks
delay in the ground anchor construction programme
which was 7 months duration.

103

8 COMPUTER MODELLING OF ANCHORED


STRUCTURE
The computer modelling was carried out using a
hardening soil model and a staged approach where
the excavation was modelled to each anchor
installation level with pre-stressing of anchors at that
level before the excavation progressed to the next
level. The initial modelling of deep excavation, with
two dimensional finite element analysis using Plaxis
and soil parameters summarised in Table 2,
indicated a maximum lateral wall movement of
85mm. A comparison with actual field measurements
using inclinometers showed that the actual lateral
movements of the wall to be significantly less than
this initial prediction. Further investigation into
modelling and site conditions revealed that the
dewatering system, designed to operate throughout
the excavation period, effectively reduced the pore
water down to the final excavation level. The system
was employed around the whole perimeter of the
diaphragm wall. This reappraisal of the ground water
regime was in contrast to the initial assumption by
the design team that pumps lowered the water level
step by step to within one meter below of each
anchoring platform level. Reconfiguring the model
with the actual hydrological profile in the vicinity of the
wall resulted in an increase of effective stresses in
the soil in front of the wall which in turn increased the
passive earth pressure and thereby reduced the
predicted wall movement down to a maximum of
25mm. As shown in Figure 7, this revised analysis
provides a closer approximation to the actual
behaviour of the anchored structure.
Table 2 Summary of soil parameters
Soil
classification
Sandy loam
Clay with
layers of sand
Clay with
gravel and
lenses of
sands
Sandy loam
Dense sand
Loamy sand
Fine sand
Impervious
clay

:
kN/m3
20.5

E :
MPa
20

:
deg
17

c :
kPa
50

19.7

18

16

52

21.0

38

19

82

21.2
20.5
20.2
21.1

29
48
14
45

20
37
27
38

75
3
20
6

20.2

24

24

56

Note: unit weight, E = drained Youngs Modulus, =


effective angle of shearing resistance, c = effective cohesion

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

104

The support of a 25m deep excavation in difficult ground conditions using Single Bore Multiple Anchor
technology

Figure 7. Final lateral wall movement vs Plaxis predictions.

Figure 8. Development of cloud-based anchor analysis


and testing software for tablet.

9 DATA ANALYSIS AND THE DEVELOPMENT OF


ANCHORTEST

10 CONCLUSIONS AND CLOSING REMARKS

The analysis and management of the data produced


from investigation tests and the mandatory testing of
thousands of anchors proved challenging. The
project specification stipulated that no anchors could
be accepted into the works until the results from
acceptance tests were examined and signed off by
the clients representatives and this process often
created delays. These issues paved the way for the
development of intuitive and innovative cloud-based
software, designed specifically for tablets, in order to
cope with the rigors of a field operation (AnchorTest
Ltd, 2015). At Kuntsevo, the protocols for the delivery
of data had already been established during the
development phases of the software package, so
although the final product was not used on the
project, the concept design and preliminary trial
versions were optimized as a direct result of the
works carried out at Kuntsevo (Figure 8). The
completed product is now fully enabled and available
for use on all types of ground anchor applications to
permit
paperless
real-time
analysis,
data
management, GPS technology and incorporates the
acceptance criteria for major international ground
anchor codes of practice (including BS8081:1989
and PTIDC35.1-14). This innovation provides a more
reliable and convenient way to process data, both in
the field and in the office, and effectively moves
anchor testing beyond the limits imposed by data
input using spreadsheets on laptops and PCs.

The construction of the deep basement structure for


the Kuntsevo Plaza development saw the first use of
SBMA technology in Russia. Under difficult
geotechnical and environmental conditions some
3600 anchors were successfully installed stressed
and tested. The success of this project was largely
based on the decision to undertake an extensive
programme of investigation tests which included the
use of post-grouting trials. These trials provided an
opportunity for the contractor to become proficient
with the use of the ground anchor technology and
generated useful parameters to permit effective
execution of the post-grouting operations when they
were employed.
Despite the restrictions imposed by the close
proximity of the anchored structure to sensitive
services, a well-executed wall design restricted
lateral wall displacements to a maximum of 7.5mm
during the anchor installation phase. The detailed
evaluation of the construction process and the
modelling of soil properties incorporating the effects
of dewatering and the consequential increase in
effective stresses and passive resistance ensured
more credible predictions of lateral displacement
using Plaxis. The most important benefit of the
limited wall movements was the significant reduction
in settlement or disturbance to the adjacent services.
The conservatism in the anchored wall design
generated considerable savings by avoiding top
down construction, proposed compensation grouting
and proposed realignment of an existing sewage
pipe.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

MOTHERSILLE D. et al.

The project provided the environment for the


creation of an innovative cloud-based data
management and analysis tool for anchor testing
which should prove highly beneficial for the industry.
The successful execution of the SBMAs at
Kuntsevo has set a valuable precedent and paved
the way for further construction of deep basements in
soils that historically were only able to sustain relative
low working loads (Figure 9).

105

Drilling and American Society of Civil Engineers,


Geotechnical Special Publication, 124, Orlando,
FL, USA, 361373.
Mothersille DKV, Orkun AO and Okumusoglu B
(2012). The performance of Single Bore Multiple
Anchor trials installed in mixed Moscow soils,
Proceedings
of
the
Road
Construction
Conference, Penza, Russia, 461470.
Ostermayer H (1974). Construction, Carrying
Behaviour and Creep Characteristics of Ground
Anchors, Conference on Diaphragm Walls and
Anchorages, London, UK, 141151.
Ostermayer H (1977). Detailed design of
anchorages, Review of Diaphragm Walls, I.C.E.,
London, UK, 5561.
Ostermayer H and Barley AD (2003). Fixed anchor
design - Ground Anchors, Geotechnical
Engineering Handbook, Vol. 2, Pub Ernst and
Sohn, Berlin, Germany, 189205.
VSN 506-88 (1989). Design of Arrangement and
Installation of Ground Anchors, USSR Ministry of
Assembly and Special Constructions, USSR
Ministry of Assembly and Special Constructions,
Moscow, Russia.

Figure 9. Completed wall showing 25m deep excavation


and six levels of SBMAs providing effective support.

11 ACKNOWLEDGEMENTS
The authors wish to thank the site staff at ENKA and
Kasktas for their efforts in executing the works during
difficult climatic conditions. The detailed Plaxis
analysis was carried out by Mr enol Adatepe from
Kasktas A., Turkey and detailed reports on the
anchor tests were prepared by Mr Ali Orkun Bayur
during his time with Kasktas A., Turkey, and their
contributions are acknowledged with thanks. Special
thanks are also due to Dr Rasin Duzceer and Mr Alp
Gokalp of Kasktas A., Turkey for their valued
contribution to this work and the success of the
project.
12 REFERENCES
AnchorTest Ltd (2015). AnchorTest for iPad
Software http://www.anchortest.info, London, UK
Barley AD (1995). Theory and Practice of the Single
Bore Multiple Anchor System, International
Symposium on Anchors in Theory and Practice,
Salzburg, Austria, 293-301
BS8081:1989. British Standard Code of practice for
Ground Anchorages, BSI, London, UK.
BS EN1537:1999. Execution of special geotechnical
work. Ground anchors, BSI, London, UK.
BS EN1537:2013. Execution of special geotechnical
work. Ground anchors, BSI, London, UK.
Bruce ME, Traylor RP, Barley AD, Bruce DA and
Gomez J (2004). Post grouted single bore
multiple anchors at Hodenpyl Dam, Michigan,
ADSC: International Association of Foundation
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

The use of MSE walls backfilled with Lightweight Cellular Concrete in soft
ground seismic areas
El uso de muros MSE rellenados con concreto celular ligero en reas ssmicas de terrenos blandos
Daniel PRADEL 1&2 and Binod TIWARI3
1Adjunct

Associate Professor, University of California Los Angeles (UCLA), Los Angeles, California (dpradel@ucla.edu) &
2Vice-President, Shannon & Wilson, 664 W. Broadway, Glendale, California (dep@shanwil.com)
3Professor, California State University Fullerton (CSUF), Fullerton California (btiwari@csuf.edu)

ABSTRACT: A series of numerical analyses were performed on a Mechanically Stabilized Earth (MSE) wall that used
Lightweight Cellular Concrete (LCC) instead of soil as infill. These analyses were performed using the geometry and input
ground motions for a wall recently built for the Silicon Valley Rapid Transit (SVRT) system near San Francisco, California.
For our analyses, the LCC-MSE wall was significantly weakened in our numerical models by using shortened geogrid
lengths, and lower material strengths than the constructed wall. In spite of the weakened nature of the wall analyzed herein,
seismic failure of the LCC materials and supporting ground was not predicted. Our analyses show that well designed LCCMSE walls tend to move dynamically in a quasi-rigid fashion, i.e., that they move mainly laterally and do not exhibit major
rocking or seismic settlements. Additionally, because of their broad base of MSE walls, these walls distribute compressive
and shear stresses to the underlying ground in a relatively even manner. Our numerical analyses also show that internal
reinforcement of LCC-MSE walls is important to restrain side panels during earthquakes, but that the inertial loads from the
panels are quickly transferred to the LCC. Hence, that long or continuous reinforcements are not needed for seismic stability.
In summary, our analyses show that LCC is an excellent material for MSE walls and that the lightening of vertical loads that
LCC provides has distinct seismic advantages in soft ground seismic areas, e.g., the elimination of ground improvement.

1 INTRODUCTION
Soft-ground
construction
poses
significant
geotechnical challenges, ranging from large
consolidation settlements (below the structure and in
nearby developments), construction staging and
extended project schedules. In seismic regions, soft
ground conditions often result in the significant
amplification of structural demands. For freeway and
railroad embankments, such demands often result in
costly ground improvement to mitigate the significant
consolidation settlements resulting from the heavy
weight of Mechanically Stabilized Earth (MSE) walls.
Recently, a novel approach for the construction of
MSE walls has been used, which involves replacing
the MSEs soil infill with Lightweight Cellular Concrete
(LCC). The main advantage of LCC is its low unit
weight (often about half the unit weight of water). In
California, examples of LCC-MSE walls include the
Colton Crossing for the Union Pacific-BNSF railroad
in Colton (Teig and Anderson, 2012), the San Bruno
Railroad Grade Separation in San Bruno, and the
Silicon Valley Berryessa Extension in San Jose which
will connect the Silicon Valley Rapid Transit (SVRT)
system to San Franciscos Bay Area Rapid Transit
(BART) system.

In addition to railroad projects, LCC-MSE walls


have also been used and/or are being considered for
road transportation projects such as the 22/405
freeway interchange in Orange County, California, as
well as for various new bridge approaches.

Figure 1. Lightweight Cellular Concrete.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

108

The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground seismic areas

Figure 2. Construction of an LCC- MSE wall (Cell-Crete,


2015).

One of the main advantages of LCC is a reduction


of up to 75% in the static weight of a traditional MSE
walls. This significant weight reduction, results in
much lower consolidation settlements, as well as
reduced inertial loads and dynamic compression
during earthquakes.
The excavation of relatively heavy on-site soils for
the wall foundations, combined with the low unit
weight of LCC (compared to soil) can result in a
balanced design (with zero added bearing pressures)
or a small net increase in the foundation soils vertical
stresses. This has allowed designers to either
completely eliminate the need for improving the soft
subgrade (e.g., eliminating stone columns) under
certain MSE walls or at least to significantly reduce
the scope of ground improvement (e.g., limited vibroreplacement in Teig and Anderson, 2012).
An advantage of using LCC as fill, is that the
relatively high strength of this material (compared to
conventional MSE granular fill) also results in
essentially no lateral earth pressures on the MSE
panels and abutment walls.
The weight advantages of LCC are shared by other
materials, such as Geofoam. However, Geofoam is
combustible and reacts chemically with hydrocarbons
such as diesel fuel. Hence, railroads have been
reluctant to accept Geofoam structures in California,
but have often accepted the use of LCC.

low thermal conductivities, and high strengths


compared to that of a conventional MSE soil infill. Its
nature and behavior is often described as similar to
that of porous soft rock.
In California, the design engineer generally
specifies a minimum compressive strength for the
LCC material that the supplier (or manufacturer) must
deliver during construction. It is not unusual to have
two or more minimum strengths specified by the
design engineer for the same wall, e.g., a higher
strength near the foundation soils and lower strengths
in the upper portions of the wall. In our experience,
densities of LCC used for MSE walls in California
typically range from about 300 kg/m 3 to about
600 kg/m3. The strength of LCC correlates strongly
with the density of the material; for the above range of
densities, the average uniaxial compressive strengths
of LCC materials, of the type used in MSE walls in
California are typically between 500 and 3,000 kPa.
Construction of LCC-MSE walls
Except for the foundations, the LCC in MSE walls is
generally formed between the facing panels
(Figures 2 to 5). These panels play an important
protective role as LCC is not as strong as conventional
concrete and can be relatively easily damaged, e.g.,
by a small vehicle impact. The facing panels are
anchored to the wall through reinforcements, as in
conventional MSE walls.
In our experience, LCC-MSE walls have typically
been reinforced with metal straps, steel rods and
geogrids (Figures 3 to 5). The reinforcement of the
LCC in MSE walls is sometimes continuous (as in
Figure 3), or limited in length to the area near the outer
face (as shown in Figure 4).

2 CONSTRUCTION OF LCC-MSE WALLS


Nature and Physical Properties of LCC
Lightweight Cellular Concrete (ACI, 2006) is created
by adding stable air cells during the manufacture of
the material (Figure 1), and LCC is placed in a manner
quite similar to concrete (Figure 2). Relatively few
MSE walls with LCC as infill have been built in seismic
areas and its use is generally considered a novel
practice.
The manufactured process of LCC results in a
concrete preformed foam that has very low densities,

Figure 3. Placement of LCC over continuous geogrid


reinforcement, within the panels of an MSE wall being built
for the SVRT system near San Jose, California (design of
wall detailed in GDC, 2014).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

PRADEL D. et al.

Figure 4. LCC placement in an MSE wall reinforced with


metal straps near the facing panels (Cell-Crete, 2015).

109

Figure 7. Detail of FLAC model including geometry of LCC


and concrete slab. Stage shown is before backfilling next to
wall (each underlying grid is 3.3 m = 10 feet).

3 NUMERICAL MODELING
Introduction

Figure 5. LCC placement by in an MSE wall reinforced with


steel rods (Cell-Crete, 2015).

Figure 6. FLAC model including geometry of LCC and


subsurface soft soil profile (each underlying grid is 3.3 m =
10 feet).

To understand the seismic behavior of a geogrid


reinforced LCC-MSE wall we conducted numerical
analyses using the computer program FLAC
version 7.0 from Itasca (2011). For the analyses
shown herein, we adopted the subsurface conditions,
wall geometry and seismic design loads of an LCCMSE wall that was recently completed in 2014-2015
for the Silicon Valley Rapid Transit (SVRT) in San
Jose, California (shown in Figure 3). This project will
extend subway service from San Francisco to San
Jose.
The adopted model geometry for the LCC-MSE
wall is shown in Figures 6 and 7. It includes about 9 m
of LCC materials built in 8 stages (about 2 m below
and 7 m above final grade), and a rolling slab about
1.5 m thick at the top composed of conventional
reinforced concrete.
Since, in our opinion, the SVRT wall (being built in
Figure 3) was well designed but very conservative, we
decided to analyze a weakened version of the
constructed wall. As a result, certain aspects of our
numerical analyses are significantly different than
those used for the original design by the design
engineers as well as the final construction. For
example: (a) the geotechnical consultants numerical
analyses did not include geogrid elements in their
FLAC models (GDC, 2014), (b) the final construction
incorporated full length geogrid between the panels
(as shown in Figure 3), (c) in our analyses geogrid
reinforcement only extends 1/3 of the length between
the walls, (d) we used reduced (lower values) LCC
strengths than specified during construction, and (e)
we used in FLAC the dynamic properties obtained
from cyclic simple shear tests recently performed at
California State University Fullerton (Tiwari, 2015a &
2015b) for the Cell-Crete corporation (Cell-Crete,
2015). Please note that Professor Tiwaris triaxial and
dynamic simple shear tests were performed on a large
number of LCC samples that were prepared for

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

110

The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground seismic areas

different sets of LCC densities, and that we used only


the data that was the most appropriate for the adopted
LCC density in our analyses, i.e., 480 kg/m3.
In summary, the analyses presented herein, should
be considered applicable to a hypothetical MSE wall
that has a geometry similar to the one shown in
Figure 3, but that is purposely much weaker. The
differences between the constructed and analyzed
wall are considered substantial.
Adopted Properties of concrete and LCC
Materials
Both the heavy reinforced concrete slab and LCC
materials were modeled as elastoplastic materials
with a compressive strength of 20,000 kPa and
275 kPa, respectively. Please note that the adopted
LCC strength is conservative for the specified density
of 480 kg/m3. As previously indicated, our adopted
LCC compressive strengths are significantly lower
than the strength specified for the constructed wall in
Figure 3, where we understand minimum LCC
strengths around 550 kPa were ultimately required
and vastly exceeded by the LCC supplier (Cell-Crete).
Dynamic Properties of soils and LCC
Materials
The shear wave velocities near subject site and in San
Jose, California, are known to considerable depths
and have been the subject of numerous studies.
Typical shear wave velocity profiles are reported in
Chiu et al. (2008) for depths of 0 to 410 m (including
the deep suspension logging performed by the
USGS).
For our numerical analyses, we used both the
specific shear wave velocities obtained from the sites
subsurface characterization (GDC, 2014) as well as
the deeper data in Chiu et al. (2008). The adopted
shear wave velocities varied linearly in clays
from 150 m/s at the surface to 300 m/s a depth of
30 m. Sand layers below 30 m were modeled with a
constant 450 m/s velocity. Similarly, undrained shear
strength varied from about 75 kPa near the surface to
about 160 kPa at a depth of 30 m.

Figure 8. Accumulation of permanent shear strains and


hysteretic damping in FLAC based on Massing (1926) rule.

Figure 9. Matching of modulus reduction and damping


curves used for clays and Vucetic & Dobry experiments
(1991).

To allow damping to vary with time during dynamic


loading in FLAC, we adopted, for simplicity, hysteretic
damping based on the Massing (1926) rule and we
used as backbone the shear moduli curves at specific
depths. This technique allows the accumulation of
shear strains during dynamic loading as shown in
Figure 8. Please note that a small amount of Rayleigh
damping of 0.2% was added for numerical stability as
well as energy dissipation at small loading cycles.

Figure 10. Matching of modulus reduction and damping


curves used for clays and Darendeli experiments (2001).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

PRADEL D. et al.

The modulus reduction curves that we used in this


study were based on Vucetic & Dobry (1991) for the
clay materials (i.e., in the upper 30 m) and Darendeli
(2001) for the sands (between 30 m and the base of
our model at 45 m). Since FLAC uses mathematical
expressions instead of curves, we matched the
experimental curves by Vucetic and Dobry as well as
Darendeli as close as possible in the main area of
concern, i.e., for shear strains of 0.05% to 0.2% (as
shown in Figures 9 and 10). Note that the damping
shown in Figures 9 and 10 is directly obtained from
the modulus reduction curves through the use of the
Massing (1926) rule.
Geogrid, rolling slab and panels
The adopted MSE reinforcement in our analyses
consisted of Tensar UX1400 geogrid, which is one of
the weakest uniaxial geogrids in the UX series. This
geogrid has a tensile strength of 27 kN/m at 5% strain.
The geogrids were modeled in FLAC using axial
elements (known in FLAC as cable elements).
Due to its thickness the concrete rolling slab, which
seats on top of the LCC-MSE wall, was modeled using
solid elements. To account for the inertial loads due to
ballast, rails and similarly related machinery and
equipment, the unit weight of these solid elements
was increased appropriately.
The MSE walls side panels were modeled in FLAC
using liner beam elements having a thickness of 3 cm.
These elements were given the properties of
reinforced concrete and were directly attached to the
end of the geogrid layers.
Figure 11 depicts the computed vertical stresses
and tensile forces in the geogrid layer at the end of
construction. Please note that the predicted tensile
forces in the geogrid layers are less than 2% of its
tensile strength. These tensile values are low due to
the relatively high shear strength of the LCC materials.

111

4 GROUND MOTIONS AND NUMERICAL


ANALYSES
To minimize wave reflections at model boundaries a
quiet (viscous) boundary was specified along the base
of the model, and free-field boundaries were specified
along the edges, as shown in Figure 12.
The ground motions used in our analyses were
obtained from existing records of earthquakes with
magnitudes and accelerations similar to those
anticipated at the San Jose site (Abrahamson, 2012)
and included both Fault Parallel and Fault Normal
components for each of the records.
The motions were spectrally matched (to the design
spectral accelerations shown in Figure 13) using the
software codes RspMatch and RspMatchEDT, pre
and post processors for RspMatch (Abrahamson,
1992, and Geomotion, 2011a). The surface ground
accelerations and spectral accelerations (before and
after spectral matching) are shown in Figure 13.

Figure 12. Quiet and free field boundaries used in FLAC.

Surface ground motions were deconvoluted in


order to obtain the input velocities at the base of our
FLAC model using the program SHAKE2000
(Geomotions, 2011b).
The results of our numerical simulations are shown
in Figures 14 and 15. During our simulations, the
computed maximum horizontal displacements ranged
from 1 to 4 cm, and we predicted differential vertical
movements from to 1 cm.
Figures 14 and 15 show the deformed mesh in an
exaggerated scale from the beginning (top plot) to the
end of the earthquake (bottom plot). As can be seen
the mesh moves mainly in a horizontal and quasi-rigid
manner. Rocking of the LCC-MSE wall is very minor
and not noticeable in Figures 14 and 15.

Figure 11. Computed geogrid tensile forces and vertical


stresses in ground, LCC wall, and rolling slab.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

112

The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground seismic areas

(shear, bearing, etc.). Similarly, our numerical


analyses did not predict failure in the walls LCC
materials either, i.e., we did not observe tensile, shear
or compressive failures of the LCC infill.
As seen in Figures 14 and 15, the side panels have
inertial loads that pull on the geogrid layers and result
in an increase of geogrid tensile forces. Figures 14
and 15 show that these tensile loads are almost
immediately transferred to the LCC during an
earthquake, and it appears that after approximately
2 m the geogrid has virtually no beneficial seismic
role. Hence, the main role of geogrids appears to be
for crack control purposes (e.g., cracks resulting from
material shrinkage or minor differential movements
resulting from varying bearing conditions).
5 CONCLUSIONS

Figure 13. Accelerations (in g) of the 1992 Landers


earthquake and spectral accelerations, before and after
spectral matching.

Because the reinforced concrete rolling slab at the


top of the wall is heavy, it does create inertial
overturning moments. However, these moments
appear to be easily countered by the large base of the
LCC-MSE wall which distributes them relatively
evenly and prevents large vertical strains in the
subgrade soils near the edges of the MSE wall.
Even during the most intense portion of the
earthquake (mid-plots in Figures 14 and 15), the
vertical and horizontal stresses are relatively evenly
distributed throughout the base of the wall, and do not
lead to a shearing along the base or bearing failures
below the LCC-MSE wall. In summary, our model
does not predict a soft ground failure of any type

Our numerical simulations were performed on a


significantly weaker version of the LCC-MSE wall
shown under construction in Figure 3, that was built at
a soft clay site in the City of San Jose, for the Silicon
Valley Rapid Transit system near San Francisco,
California. The designers (GDC, 2014) used LCC for
this wall in order to significantly lighten the load of an
originally proposed traditional MSE wall which used
soil as infill. The switch from soil to LCC was very
successful and allowed the designers to completely
eliminate the need for soil improvement under this
MSE wall.
Our FLAC analyses indicate that on soft clay sites,
LCC-MSE walls, such as the one analyzed herein,
move in a quasi-rigid manner during earthquakes, i.e.,
they move mainly parallel to the ground surface and
do not develop significant total or differential
permanent vertical seismic movements.
Our numerical modeling which was conducted on a
weaken version of a constructed LCC-MSE wall, did
not predict ground failures (shear, bearing, sliding,
overturning, etc.) under static (during construction) or
under design seismic conditions. Similarly, failure of
the LCC materials in either compressive, tensile or
shear modes of failure was not predicted.
Our analyses indicate that the role of the geogrid
reinforcement during earthquake loading is mainly to
hold the side panels and that for seismic loading
geogrid reinforcements provide little benefit to the wall
beyond a distance of about 2 m from the face of the
wall.
Hence, the main role of geogrids in LCC walls appears
to be for crack control purposes (e.g., cracks resulting
from material shrinkage and/or minor differential
movements
resulting
from
varying
bearing
conditions), and designers may consider limiting both
the reinforcement lengths and type.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

PRADEL D. et al.

Figure 14. Deformed mesh, vertical stress contours and


geogrid tensile loads (from beginning to end of earthquake).

113

Figure 15. Deformed mesh, horizontal stress contours and


geogrid tensile loads (from beginning to end of earthquake).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

114

The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground seismic areas

REFERENCES
Abrahamson, N.A. (1992). Non-stationary spectral
matching, Seismological Research Letters 63(1),
30.
Abrahamson, N. (2012) Updated Peer Review of the
2005 ARS Curves for the Silicon Valley Rapid
Transit Project, dated August 13, 2012.
ACI (2006). Guide for Cast-in-Place Low Density
Cellular Concrete, ACI-523.1 R-06.
Cell-Crete (2015), Cellular Concrete Engineered Fill
from Cell-Crete, www.cell-crete.com.
Chiu P., Pradel D. Et al. (2008), Seismic Response
Analyses for the Silicon Valley Rapid Transit
Project,
ASCE
GSP
181,
Geotechnical
Earthquake Engineering and Soil Dynamics IV,
Sacramento, 1-10.
Darendeli (2001). Development of a new family of
normalized modulus reduction and material
damping curves, Ph.D. Dissertation, Dept. of Civil
Engineering, Univ. of Texas, Austin, TX.
GDC (2014), Dynamic Numerical Analyses using
FLAC for Lightweight Cellular Concrete MSE Walls
for BART Berryessa Extension, San Jose, CA,
report by Group Delta Consultants dated July 11,
2014.
Geomotions (2011a), RspMatchEDT user manual,
www.geomotions.com/Download/RspMatchEDTM
anual.pdf.
Geomotions (2011b), SHAKE2000, user manual.,
www.geomotions.com/Download/SHAKE2000Man
ual.pdf.
Itasca (2011), FLAC (Fast Lagrangian Analysis of
Continua) Version 7.0, Minneapolis, USA,
www.itascacg.com.
Masing, G. (1926), Eigenspannungen and
verfertigung beim messing, Proc. 2nd Int.
Congress on Applied Mechanics, Zurich.
Teig J. and Anderson J. (2012). Innovative Design for
the Colton Flyover Grade Separation of IPRR and
BNSF, Colton, CA AREMA Annual Conference &
Exposition.
Tiwari B. (2015a), Preliminary Result Static Shear
Strength of Lightweight Cellular Concrete Sample
Batch 1, report dated January 26.
Tiwari B. (2015b), G over Gmax & damping tests,
personal communication of April 4th.
Vucetic & Dobry (1991), Effect of Soil Plasticity on
Cyclic Response, ASCE Journal of Geotechnical
Engineering 117(1), 89-107.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

3rd.

International Conference on Deep Foundations


Deep Foundations and Soil Improvement in Soft Soils

Session 3:
Soil improvement

Technical Committee

TC-214

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Soil improvement around the world Applications and solution examples


Mejoramiento de suelos alrededor del mundo Aplicaciones y ejemplos de solucin
F. GERRESSEN1
1

BAUER Maschinen GmbH, Schrobenhausen, Germany

ABSTRACT: The uses of soil improvement to utilize areas, which are not suitable for foundation purpose, have a long
tradition. Due to the increasing demand of foundation works in areas of unsuitable soils, the improvement of the existing
soils becomes even more important in the future. Various applications and systems for various demands exist around the
world. Either it is a simple improvement for settlement reduction or a more important aim as liquefaction mitigation.
However, each system has its advantages, limits and needs. The paper will describe some basics of various techniques
like e.g. Vibro-flotation, Dynamic compaction and soil mixing. The main part will focus on jobsite examples, where the
systems provided solutions for different purposes, e.g. simple foundation or liquefaction mitigation.

1 GENERAL
Soil improvement techniques continue to progress in
addressing ground engineering problems across the
world, especially in urban areas where land
development and reuse need to be efficient not only
in geotechnical engineering but in time, cost and
energy used. These techniques provide a toolbox for
geotechnical engineers who look for opportunities to
modify the ground characteristics and behavior, not
accepting the soil as it is. Focus can be taken as well
on geo-mechanical properties like e.g. strength and
deformability as on the hydraulic conductivity, which
might be either increased or decreased. A big focus
can also be given to the aim of liquefaction
mitigation. Based on the analysis of the local
conditions, potentials and limitations of a possible soil
improvement one has to do a pre-selection of
possible techniques out of a wide range of
applications. Looking to the literature, all techniques
can somehow be put into different categories,
whereby the definition of the categories is not strict.
For instance, once you find the vibro-flotation in the
category of vibro and impact compaction, which is
more related to the soil conditions and the way of
improvement, another time you find it at the deep
vibro techniques, which is more related to the used
equipment.
2 TECHNIQUES
Trying to categorize the various techniques in a way
that the number of categories will not be too high, will
lead to a minimum of five categories.

- Self-compaction by Vibro or impact compaction


- Dewatering by vertical drains or Vacuum
consolidation
- Soil Mixing
- Soil-displacement by rigid inclusions or stone
columns
- Grouting
A classification of ground improvement techniques
adopted by the TC 211 (Former TC17) of the
International Society for Soil Mechanics and
Geotechnical Engineering (ISSMG) covers 29
different methodologies also in 5 different categories.
Some of the most important techniques will be
described in the following.
2.1 Vibro compaction, also known as vibroflotation (VF)
is one of the oldest soil improvement techniques,
developed back in the 1930th by the Keller Company.
VF is applicable in non-cohesive and slightly
cohesive granular soils such as sands and gravels,
as well as slag deposits.
It is suitable for carrying high loads on the
improved subsoil, including dynamic loads without
significant settlements. It offers a particularly
economical application in fully saturated soils below
the groundwater table.
The hydraulically (or electrically)-powered vibrator
assemblies will be suspended from crawler-mounted
cranes. The power will be supplied by either the
enhanced hydraulics of the special cranes or by a
power-pack mounted on the back of the crawler
crane.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

118

Soil improvement around the world Applications and solution examples

Figure 1: Crawler crane mounted depth vibrator left picture without, right picture with power pack.

The flushing medium can either be water taken


from rivers or existing groundwater. Both fresh and
salt water are suitable. The flushing volume ranges
typically from 50 to 90 m/h. It is pumped at a
pressure of about 6 to 8 bar to the top of the vibrator
assembly.
2.2 Dynamic compaction
is a soil densification method developed by
Menard back in the 1960th.
The soil is compacted by repeated dropping of a
heavy weight (pounder) from a predetermined height
onto the ground surface. The imparted high kinetic
energy, which is transmitted to deeper soil layers,
forces the soil particles into a denser state of
compaction.
The degree of compaction depends on the weight
of the pounder, the height from which the pounder is
dropped, and the compaction grid is dropped.
A heavy weight (pounder) is dropped in several
passes in a primary, secondary and often tertiary
grid. The primary grid (widest spacing) is used to
achieve compaction at depth. It uses the largest
weight and highest drop. The secondary and tertiary
grid is used to achieve compaction at medium and
shallow depths. The process is completed by
compacting the surface layer across the treatment
area in a final contiguous ironing pass.

Figure 2: Heavy-duty Crawler crane mounted with steel


pounder.

2.3 Soil Mixing


is a methodology were the soil is improved by
mixing it with cement, lime or other binders in situ by
using a mixing tool. This methodology can be used
either as wet mixing or as dry mixing.
In the more frequently used wet mixing process,
usually a mixture of binder and water, maybe with
additives, is injected and mixed with the soil.
Depending on the type of soil and binder, by the end
of the mixing process, a mortar like mixture is
created which hardens during the hydration process.
In the dry mixing process, the binder is directly mixed
with the soil and reacts directly with the existing soil
and water.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

GERRESSEN F.

Mixing tools show several versions, so that they


are mixing around either vertical or horizontal axis,
mixing in a trench and maybe are jet assisted. Figure
3 shows a classification done by Bruce 2010.

119

Depending on the technique, improvement depth


of up to 80 m can be achieved.

Figure 3: Updated Deep Mixing Method classification (Bruce, 2010).

Even the history of these techniques starts


predominantly in Japan (wet mixing) and Scandinavia
(dry mixing) in the 1970th, nowadays soil mixing is
getting more and more important all around the
world.
One of the latest developments in the field of wet
mixing was the Cutter Soil Mixing (CSM) method,
developed in a joint venture of BAUER Maschinen
GmbH and Bachy Soletanche about 10 years ago.
CSM is used mainly for stabilizing soft or loose soils
(non-cohesive and cohesive), however the machinery
used, derived from Bauers cutter technology,
extends the applicability of the method to much
harder strata when compared to other methods of
soil mixing.

2.4 Rigid inclusions (RI) and stone columns (SC)


are classified by the TC211 as ground
improvement methods with admixture or inclusions.
They can be installed using various techniques.
While the SC method is quite old, based on the
development of vibro compaction back in the 1930th,
the use of RI is first mentioned in the 1980th.
Usually, the construction takes place using a
regular grid of vertical elements, triangular or square,
in soil layers with low bearing capacity and/or high
compressibility, either down to a competent layer or
to a defined depth where no competent layer is
reachable. Due to the fact, that the elements are
stiffer than the surrounding soil, they take part of the
load, with the aim to provide in combination with the
surrounding soil sufficient bearing capacity and/or
acceptable settlement resistance.
There are a few differences between the RIs and
SCs.
In comparison to the SCs, RIs dont require lateral
confinement from the existing soil due to their
properties in term of stiffness/strength. Typical
methods used for RIs are e.g. the well know
Controlled Modulus Columns (CMS) or Vibro
Concrete Columns (VCC).
Replacement ratios are in a range of 10 to 35 %
for SCs, but only 2 to 10 % for RIs. Typically RIs
require a load transfer platform.

Figure 4: BCM Mixing head for CSM.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

120

Soil improvement around the world Applications and solution examples

3 JOBSITE EXAMPLES
3.1 CSM Technology at Frjus, France
For the foundation of an industrial building on a site
with up to 6 m low strength colluvial soil over a marlsandstone, a soil improvement solution had to be
found. Main purpose of the design was as well the
minimization of the total as the differential settlement
of the structure, which covers about 3600m.
The performed solution was a combination of soil
cement panels, carried out by the CSM technology,
overlaid by a load transfer layer, with a final
placement of the concrete slab of the building (Figure
5).

samples from the final CSM panels were important to


adjust execution parameters and obtain required
properties of the soil cement panels.
Samples taken to prove the achievement of the
required UCS value of 3 MPa as a minimum, design
value of 1.5 MPa UCS at working load with a safety
factor of 2, showed a good success.
The choice of CSM technology could be seen as
the most suitable solution under the local
circumstances, as it combines a number of
advantages,
technical,
economical
and
environmental. For instance the ability of executing
the system in all jobsite soil conditions, including the
Marl-Sandstone, and its use as construction material
with the benefit of spoil reduction. In addition, the
system allows an installation with reduced vibrations.
3.2 Stone Columns at Formosa Plant, Taiwan

Figure 5. Cross section and geometry for finite element


modelling.

The CSM panels were carried out as single panels


at a size of 0.6 x 2.4 m in a rectangular pattern
(Figure 6), with enlarged panel caps of 1.8 x 3.6 m to
allow more efficient load transfer. All panels needed
to key into the Marl-Sandstone for a minimum of 0.5
m.

Observing Taiwan's chronic shortage of upstream


petrochemical materials, the Formosa Plastics Group
proposed the Naphtha Cracking Project. Following
several rejections by the government, final approval
was obtained to build the No. 6 Naphtha Cracking
Project. The project became part of a huge offshore
industrial complex development with a total of 41
industrial plants. In addition to the industrial plant a
new harbor and a new town were built as well.
The development started with huge land
reclamation. About 100 Million m of marine sand
were spread out over an area of more than 200
hectares with approximately 8 km in length along the
coastline and about 4 km into the sea. Heavy
structure were founded on driven pre-cast piles, but
the remaining area was first improved by dynamic
compaction followed by stone column installation for
the deeper parts up to 20 m.
Main purpose for the decision of ground
improvement by the use of stone columns was to
reduce the liquefaction potential of the soil during
earthquakes. The earthquake design criteria
specified a magnitude M = 7.3 and an acceleration of
a = 0.21 g.
Based on the design criteria, the existing soil
parameters and the chosen equipment, a 2.5 m
triangular spacing, exemplary shown in figure 7, was
considered as feasible. Test sections were carried
out to proof and establish the right spacing and
production parameters.

Figure 6: Rectangular pattern of CSM panels with panel


cap enlargement.

Due to the use of soil as construction material,


quality control during installation by using the
operators monitor to control the installation
parameters in real time as well as the test of
achieved unconfined compressive strength on
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

GERRESSEN F.

121

Figure 8: Plant and earthquake location.


Figure 7. Typical layout for stone column arrangement.

In total 1,500,000 linear m of stone columns were


installed in a depth range of 13 to 20 m from 19972000. About 60 %, in particular the deep ones were
installed using the Bauer TR vibrator and the bottom
feed method, which ensures the forming of a
continuous stone column over the full depth.
The design could be proved already, as in 1999,
two earthquakes took place. First the Chi-Chi
earthquake with a magnitude of M = 7.3 on
September, 21st, followed by the Chia-I earthquake
with a magnitude of M = 6.8 on October, 22nd.
Especially the Chi-Chi earthquake, also known as
the 921 earthquake, caused disastrous damages.
More than 2,400 people were killed, over 50,000
building were destroyed, and another 50,000
buildings were severely damaged.

Immediately after the earthquake event all plant


operation was stopped looking for immediate survey
of the structures by consultant firms CTCI and RESI.
The survey showed no structural damage of
structures and tanks at all at a maximum settlement
of 10 mm. Therefore, the plant could get back in
operation in full scale very shortly after the
earthquake.
Also the Geotechnical Earthquake Engineering
Server (GEES), which provide constant information
about worldwide earthquake impacts states in one of
their reports: Liquefaction happened at untreated
ground at Formosa Plastics Industrial Park in Mailiao.
This park is a reclaimed island, but no liquefaction or
ground failure was noted at treated areas within this
park; Formosa Heavy Industries Corporation used a
combination of dynamic compaction, preloading and
stone columns, with piles generally supporting
buildings.

Figure 9. Damages in Tai Chung harbor September 21st.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

122

Soil improvement around the world Applications and solution examples

4 CONCLUSION
The wide range of soil improvement techniques offer
a wide range of applications to utilize soils, which are,
due to their mechanical properties, usually not
suitable for foundation purpose. One of the most
important applications can be seen in the use of
liquefaction mitigation.
REFERENCES
Wolfgang Brunner, A.rthur Bi, Yan Lian Chen 2002.
Ground improvement by stone columns at
Formosa plant Taiwan and its earthquake
response, Ninth international conference on Piling
and deep foundation, Nice, 2002.
Franz-Werner Gerressen, Thomas Vohs, 2009. Soil
Improvement, Vibroflotation, Vibroreplacement
and concrete columns, Simposio International, Las
Technologias y los Sistemas de Cimentation para
el siglo XXI, Mexico City, 2009.
Bruno Simon, General Report-Rigid Inclusions and
Stone
Columns,
2012,
ISSMGE-TC211
International Symposium IS-GI, Brussels, 2012.
Artur Peixoto, Estala Sousa, Pedro Gomes, 2012,
Solution for soil foundation improvement of an
industrial building.
Donald Bruce, Deep Mixing in the United States:
Milestones in evolution, Deep Mixing 2015, San
Francisco 2015.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Principles and application of soil mixing for ground improvement


Principios y aplicacin de la tcnica soil mixing para mejoramiento de suelo
Charles M. WILK1
1

ALLU Group Incorporated

ABSTRACT: Mass stabilization (MS) through soil mixing is a ground improvement technique that can prepare areas of low
bearing strength soil for subsequent infrastructure development. The technique involves mixing binding agents such as
portland cement, fly ash, slag, or lime into the subject soil while the soil remains in-place (insitu). The binders cement the
soil grains together to form a cement modified soil or a soil cement. The MS-treated area now has improved bearing
capacity to support infrastructure or to prevent movement of the material such as landslides. The same insitu soil mixing
technique can be used to address contaminated areas. Binders or reagents are mixed into soil. The treatment protects
human health and the environment by immobilizing hazardous constituents within the treated material. When used for the
purpose of environmental remediation the technology is called Insitu Solidification/Stabilization (ISS). Both Mass
Stabilization and ISS treatments require laboratory studies to develop a mix design of soil and binder(s) that produce the
desired physical and/or chemical properties. The mix design is then transferred into the field. Successful MS and ISS
treatments rely on reproduction at full-scale of the mix design and the thorough mixing attained at laboratory scale. Fifty to
70% of the cost of a MS or ISS project is in the cost of the binding agent that is to be mixed into the subject soil. Underdosing, overdosing, non-thorough mixing, and mixing in the wrong areas all create cost over-runs. This paper will discuss
recent innovations in mass stabilization systems that improve the cost effectiveness of the treatment technology.
Specialized equipment can impart greater mixing shear thus improving the thoroughness of mixing. Dry powder pressure
feeders can conserve the drying capacity of binder resulting in higher strengths at lower binder dosages. Global
Positioning System (GPS)-based systems can guide the mixing operator for complete mixing coverage. An integrated
tracking and feeding system can record that proper dosing and mixing was accomplished and generate construction QA/QC
reports for the client.

1 INTRODUCTION

2 PRINCIPLES OF MASS STABILIZATION

New or expanding infrastructure projects may


encounter areas of low bearing capacity soil or areas
where marginal materials such as waste, dredged
material, or sediment have been placed. Mass
stabilization is a geotechnical method that can be
used to improve the bearing capacity or stiffness of
soil. Mass stabilization involves mixing binding
agents into soil subject to treatment while the soil
remains in-place or insitu.
Mass stabilization technique can also address
contaminated soil therefore, the same technique can
be used to attain both civil and environmental
engineering goals. In this environmental application
the
technology
is
known
as
insitu
solidification/stabilization (ISS). ISS protects human
health and the environment by immobilizing
hazardous constituents within the treated material.

2.1 Properties of Mass Stabilized soil


Mass stabilization is used to improve the construction
properties of a marginal soil. Marginal soil includes,
but is not limited to, peaty soil, high water content soil,
or soil with high proportion of silt or clay. Mixing a
cementitious binder into marginal soil creates a
material similar to cement-modified soil. When
accompanied with compaction at the time of
construction and use of a higher addition rate of
cement, the mixed material may become similar to soil
cement. The goals of mass stabilization may include:
Increase in the California Bearing Ratio (CBR).
Increase in Unconfined Compressive Strength
(UCS) (Figure 1).
Reduction in plasticity characteristics as
measured by Plasticity Index (PI). (applicable
to clayey soil).
Reduction in the amount of silt and clay size
particles.
(through
agglomeration
or
cementation).

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

124

Principles and Application of Soil Mixing for Contaminated Properties

Increase in shearing strength.


Decrease in volume-change properties
(applicable to expansive soils).
Reduction in hydraulic conductivity (Figure 2).
Reduction of pore water.

to create concrete used in construction. In MS,


cement by itself without sand or aggregate, is mixed
into soil producing cement-modified soil or soil
cement. Water may be added if the soil does not
include sufficient water to hydrate the cement.
Cementitious or pozzolanic industrial by-products
such as ground granulated blast furnace slag, and fly
ash have also been used to reduce binder costs.
2.3 Bench-scale Mix Designs

Figure 1. Unconfined Compressive Strength testing.

Projects usually include some level of laboratoryscale mix design. Several addition rates of a binder
or combination of binders are mixed with
representative samples of the soil subject to
treatment. Mixing is usually done by bench-sized
power mixers such as KitchenAid or Hobart stand
mixers (Figure 3). Powered stand mixers impart
significant mixing energy and shear (Figure 4) to
speed the work and assure thorough mixing.
Thoroughness of mixing is important for
reproducibility and comparison. Mixed samples are
cured and then tested to determine if the desired
engineering/physical properties are achieved.

Figure 2 : Hydraulic Conductivity testing.

Figure 3. Bench-scale work by stand mixer.

Goals for ISS of contaminated soil may include


reduction of the leachability of hazardous
constituents, and reduction of hydraulic conductivity
thus immobilizing hazardous constituents.
An
increase in strength usually is desired as well. First
lines of paragraphs are indented 4 mm (0.16") except
for paragraphs after a heading or a blank line (Primer
Parrafo tag).
2.2 Binders
Mass stabilization involves mixing binder into soil
subject to treatment. A variety of binders have been
used including portland cement, fly ash, slag, kiln
dusts, and lime. The most commonly used binding
agent in mass stabilization is portland cement. Most
people are familiar with portland cement as the gray
powder that is mixed with sand, aggregate, and water

Figure 4. Mixing shear developed at bench-scale.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

WILK C.

125

A matrix from successful mix designs is compiled.


The matrix often includes the relative costs of the
binder(s) used in the successful mixes. Cement
addition rates may range from 60 kilograms per cubic
meter (kg/m3) to 250 kg/m3. Fine-tuning the mix
design is important from a cost perspective. Fifty to
70% of the cost of a full scale mass stabilization
project is the cost of the binder(s) used (ALLU 2015).
There is a real incentive for selection of the least
expensive mix design and efficient use of the binder(s)
that meets the desired project performance
standards.
3 INNOVATIONS IN FULL-SCALE MIXING
Figure 6. Photograph of Power Mixer and Pressure Feeder
on-site

3.1 Mixing Equipment


Mass stabilization at full scale has been done with a
variety of mixing equipment. These include deep soil
mixing augers, high pressure jets, pulver mixers or
road reclaimers, bare excavator buckets and power
mixer attachments to excavators.
Excavator-based power mixing appears to be
particularly well suited for infrastructure improvement
since: (a) the attached mixing head can reach the
required depths of treatment, (b) excavators are
commonly available, (c) excavators are easily
transported to the jobsite, (d) the range of motion of
an excavators arm reduces the amount of time lost in
moving the mixing equipment in order to treat an area
and (e) the attached power mixer imparts the mixing
shear needed for thorough mixing and efficient use of
binder(s).
Power mixers are attachments to an excavator
(Figures 5 and 6). Power mixers resemble a very
powerful rototiller on a 7-meter stem. The power
mixers are powered by the hydraulic system of the
excavator. The mixing head of the power mixer
imparts the mixing shear necessary to thoroughly mix
the binding agent into the subject soil. A nozzle
located between the mixing drums is used to inject
binder(s) at the point of mixing.

3.2 Binder Injection Wet vs Dry


Mixing binder into marginal soil requires some
method of injecting the binder into the subject soil.
Binders are injected either in wet or dry form. In wet
mixing the binder is mixed with water prior to
injection. Using portland cement as an example, the
cement is mixed with enough water to produce a
pump-able grout. This grout is then injected and
mixed into the soil. An excess amount of water is
used to produce a pump-able cement grout. Excess
because in order to make a pump-able grout a water
to cement ratio (w/c) greater than 0.5 is needed.
Portland cement chemically requires only a little less
than half its weight in water to fully hydrate- a w/c ratio
of a little less than 0.50. Highest strengths are
achieved at low w/c ratios- 0.5. (Kosmatka et al 2006)
When combined with water that may already be
present in the soil the ideal w/c ratio may be exceeded
and even more grout and corresponding cement (and
cost) may be needed to achieve the desired strength
properties of the treated soil.
Soft (marginal) areas of soil are often the result of
excessive water in the soil. Mass stabilization is very
often used to dry soil to improve its engineering
properties. Dry mixing may be appropriate in areas
where there is sufficient water already present in the
soil to hydrate the cement.
Dry mixing uses
pneumatic Pressure Feeders to feed the binder to the
augers, power mixers, or mixing equipment. Dry
mixing is often more cost effective compared to wet
mixing. Utilizing the existing water within the soil may
result in a water cement ratio closer to 0.5. Higher
strengths can be attained with lower addition rates of
cement. Note that over 50% of the cost of a mass
stabilization project is the cost of the binder.

Figure 5. Schematic of Excavator-mounted Power Mixer


with dry binder Pressure Feeder.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

126

Principles and Application of Soil Mixing for Contaminated Properties

3.3 Locating, Controlling, Recording for QA/QC


The engineering properties of mass stabilized
areas are critical for the support of buildings,
highways, roads, industrial facilities, and rail systems.
Efficient insitu mixing and meeting performance
standards requires monitoring of the actual location,
depth, binder addition, and thoroughness of mixing.
The advent of GPS-equipped mixing equipment,
computer monitored binder feed and metering
devices, and computer recording of these conditions
has made this job easier (Figure 7). Automated data
recording is used in quality control and quality
assurance documentation.

dosing and mixing was accomplished and generate


construction QA/QC reports for the client.
5 REFERENCES
ALLU OY, 2015, Mass Stabilization Manual.
Kosmatka, S. H., Kerkhoff, B., Panarese, W.C., 2006,
Design and Control of Concrete Mixtures 14th Edition,
Portland Cement Association, Skokie, Illinois.

Figure 7: Schematic of GPS and mixing data recording


system mounted on Power Mixer and Pressure Feeder.
Illustration includes excavator operators display guiding
operator- green completed blocks and red not-yetcompleted blocks and graphic of data reports.

4 CONCLUSIONS
4.1 Mass stabilization is a ground improvement
technique that can prepare areas of low bearing
strength
soil
for
subsequent
infrastructure
development. Insitu Solidification/Stabilization is a
similar technology that can address contaminated
soil. MS projects begin with bench-scale testing to
determine the effective binder and addition rate to
use. Fifty to 70% of the cost of a MS project is the
cost of the binder mixed into soil. Recent innovations
in soil mixing, binder injection, location, dosing and
recording systems improve the cost effectiveness of
MS. Specialized excavator-mounted power mixers
are capable of efficiently injecting and mixing binders
to required depths. Power mixers impart greater
mixing shear thus improving the thoroughness of
mixing. Dry powder pressure feeders can conserve
the drying capacity of binders resulting in higher
strengths at lower binder dosages. Global Positioning
System (GPS)-based systems can guide the mixing
operator for complete mixing coverage. An integrated
tracking and feeding system can record that proper
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Sustitucin dinmica aplicada en turbas de la pennsula de Yucatn


Dynamic replacement soil improvement technique applied in peaty soils in the peninsula of Yucatan
Alfredo CIRION ARANA1, Rmi CHATTE1 & Juan PAULN AGUIRRE 2
1

Menard Mxico
CIMESA, Soletanche-Bachy, Mxico

RESUMEN: Se describe el procedimiento constructivo de Sustitucin Dinmica utilizado como sistema de mejoramiento
masivo de suelos con propiedades mecnicas pobres. Se explica el concepto de mejoramiento con este tipo de
columnas granulares de gran dimetro, se enumeran las bases de su diseo, en la descripcin de la secuencia
constructiva se destacan los controles durante la ejecucin para garantizar su calidad. Se da un ejemplo real de la
aplicacin de esta tcnica de mejoramiento de suelos en suelos orgnicos tipo turba.

1 ANTECEDENTES
1.1 Descripcin del proyecto base
El proyecto en general consiste en un desarrollo
habitacional de ms de 130 viviendas a construirse
en un terreno de casi 7 ha de superficie, que es
colindante a una laguna artificial creada para poder
tener conexin con el mar caribe. El acceso
vehicular a los lotes donde se construirn las casas
se realizar a travs de vas de circulacin
clasificadas como primarias y secundarias, segn su
ancho de calzada y nivel de rasante, las cuales
llevarn a los autos directamente a los
estacionamientos de las casas ubicados en
semistanos. Debido a las propiedades mecnicas
pobres de los suelos que se encuentran en el
terreno, que son predominantemente suelos de
origen orgnico turbas de hasta 5.0 m de espesor,
el proyecto original desarrollado para construir estas
vas consista en la construccin de un viaducto
formado por marcos estructurales de concreto
reforzado prefabricado formados por pilotes
hincados hasta la roca subyacente a las turbas y
trabes de concreto, que soportaran una losa
tambin de concreto reforzado que sera la que
finalmente funcionara como el arroyo para el trnsito
vehicular y tambin como el medio para soportar las
instalaciones necesarias para la urbanizacin del
desarrollo habitacional: instalaciones elctricas,
hidrulicas, pluviales y sanitarias.

1.2 Descripcin de la propuesta alternativa


Debido a la complejidad y el costo asociado a la
construccin de las vialidades con el sistema

estructural descrito en el inciso anterior, Menard


Mxico propuso una opcin alternativa para la
construccin de stas, la cual consisti en mejorar
las caractersticas mecnicas del suelo turboso
existente mediante la tcnica conocida como
Sustitucin Dinmica, de este modo dar la capacidad
de carga necesaria y el adecuado comportamiento
en deformacin, a dicho terreno, para poder construir
como terraceras convencionales la vialidades segn
el nivel requerido por proyecto, bajo este esquema
las instalaciones se construyeron de manera
tradicional, dejando las tuberas, pozos y registros
enterrados en el terrapln final. Con esta tcnica de
mejoramiento los asentamientos totales finales son
controlados a corto y largo plazo quedando dentro de
lmites aceptables en condiciones de servicio
(menores a 2.5 cm a largo plazo) y los
asentamientos diferenciales a valores permisibles
para conservar el bombeo de las vialidades.
La Sustitucin Dinmica es una tcnica de
mejoramiento masivo de suelos que consiste en la
formacin de columnas de gran dimetro (de 2 a 3
m) formadas con material granular compactado,
aplicadas a terrenos con propiedades mecnicas
pobres, distribuidas en mallas regulares, de forma
general la masa tiene una geometra especial
punzonante y su peso varia entre 10 y 15 t. Las
alturas de cada tpicas son entre 5 y 15 m. La
formacin de estas columnas, tambin conocidas
como plots, se realiza mediante la incorporacin
secuencial y compactacin de material granular en el
terreno original a travs de la aplicacin de impactos
generados con una masa que se deja caer
repetidamente desde cierta altura con la ayuda de

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

128

Dynamic replacement soil improvement technique applied in peaty soils in the peninsula of Yucatan

una gra. La Figura 1 muestra la secuencia de


formacin de las columnas.

E; mdulo de Young
c'; cohesin drenada
; friccin drenada
cu; cohesin no drenada
El Nivel de Aguas Freticas se encontr a 1.0 m
de profundidad promedio.
3 BASES DE DISEO

Figura 1. Secuencia de formacin de columnas de


material granular plots, mediante Sustitucin Dinmica.

2 DESCRIPCIN GEOTCNICA
El predio se encuentra en la pennsula de Yucatn,
en la zona costera, prcticamente colindando con el
mar caribe. La zona anteriormente estaba ocupada
por manglares, formada por depsitos sedimentarios
de arenas, limos, arcillas y turbas, seguidos de
estrados de roca caliza.
Segn los estudios de mecnica de suelos que se
tenan, todo pareca indicar que las condiciones del
sitio eran errticas, es decir, que exista gran
variabilidad en cuanto a los espesores y tipo de
materiales. La posterior verificacin a travs de calas
realizadas en sitio en zonas especficas durante la
ejecucin de los trabajos, confirm dicha situacin.
En las Tablas 1 y 2 se presenta un resumen de las
condiciones y propiedades estratigrficas del sitio.

Dadas las condiciones disimilares de suelos que se


presentan a lo largo y ancho del sitio de
construccin, y segn la variabilidad de las
pendientes y geometra de los terraplenes finales,
segn el proyecto y las necesidades arquitectnicas
y de ingeniera, se llevaron a cabo numerosos
anlisis paramtricos y de sensibilidad para definir la
factibilidad de la solucin, as como el
espaciamiento, profundidad y dimetro de las
columnas, y los caractersticas asociadas a las
columnas terminadas.
Como parte de estos estudios, se realizaron
clculos analticos y modelos basados en el Mtodo
de Elementos Finitos, obteniendo resultados de
asentamientos totales y diferenciales crticos para las
condiciones naturales antes del mejoramiento de
suelos

Tabla 1. Condiciones estratigrficas del sitio.


Zi
(m)

U00

U01

U02
U03

+3.8
a
+1.5
+1.0
a
+0.5
-0.8
a
-1.3
>-2.4

Zf
(m)

H
(m)

Descripcin

+1.0
a
+0.5
-0.8
a
-1.3
-2.4
a
-5.8
-20.0

0.5
a
2.0

Relleno (terrapln)

1.8

Arena tipo SASCAB

1.2
a
5.0
>17.0

Turba
Roca Caliza

Tabla 2. Propiedades estratigrficas del sitio.


Unid

U00
U01
U02
U03

c'

'

cu

(kN/m)

(MPa)

(kPa)

()

(kPa)

22
18
17
24

50
30
1
500

5
2
100

38
30
15
40

Figura 2. Modelo axisimtrico con elementos finitos


realizado para una columna de sustitucin dinmica
(CSD), PLAXIS.

15
-

Donde:
Zi; Nivel inicial
Zf; Nivel final
H; Espesor del estrato
; Peso Volumtrico
SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

CIRION A. et al.

f.

129

La vialidad quedar terminada una vez que se


coloquen las instalaciones dentro del terrapln
(excavando donde sea necesario) y se
construya la carpeta de rodamiento, banquetas
y pasillos peatonales.

Figura 3. Modelo plano de deformacin con elementos


finitos realizado para el anlisis de estabilidad y
verificacin del tratamiento propuesto, PLAXIS.

4 DESCRIPCIN DEL PROCESO


CONSTRUCTIVO
Para el mejoramiento de suelos y la construccin de
los terraplenes fueron consideradas las etapas
siguientes:
a. Una vez realizado el desmonte y despalme del
terreno, inici la construccin de los terraplenes
mediante la colocacin de una plataforma de
trabajo de material granular. Esta plataforma
tuvo el objetivo de dar soporte a los equipos de
construccin de las CSD, y a la vez garantizar
su estabilidad. Fue una plataforma horizontal,
plana y drenada.

Figura 4. Aportacin de material granular para la


conformacin de CSD.

b. Con la plataforma terminada, se realiz el trazo


de las CSD en mallas regulares de 4.0 x 4.0 m,
4.5 x 4.5 m y 5.0 x 5.0 m.
c. Con una pre-excavacin en el punto de
ubicacin de una columna, inici la aportacin
de la grava y sta se comenz a compactar
mediante una masa metlica de geometra
especial que se dej caer, en cada libre, con
una gra. Cada impacto logr la insercin del
material de aportacin en el suelo natural. La
columna se form poco a poco hasta alcanzar
el criterio de paro definido previo al inicio de los
trabajos, el cual garantizaba la correcta
formacin de la columna.

Figura 5. Izado de la masa punzonante con la gra.

d. Realizada la construccin de una CSD, se


procedi a repetir el procedimiento para las
CSD subsecuentes.
e. Posteriormente se coloc una capa de geotextil
sobre la plataforma y se realiz la construccin
del terrapln final en capas de 20 cm de
espesor ver Figura 6, compactadas al 99 %
de su PVSM, hasta alcanzar el nivel de rasante
de proyecto.

Figura 6. Construccin de terrapln final.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

130

Dynamic replacement soil improvement technique applied in peaty soils in the peninsula of Yucatan

5 CONTROLES Y VERIFICACIONES DE
EJECUCIN
Una parte esencial para asegurar la calidad de los
trabajos del tratamiento de suelos aplicado es el
control de la ejecucin de las columnas y su
posterior verificacin.
Adems del control de la construccin llevada a
cabo a travs de la generacin de los registros de
construccin de cada una de las CSD, tambin fue
necesario definir criterios de paro, para tener los
parmetros de referencia necesarios que indicaran la
finalizacin de la construccin de cada CSD. Estos
criterios se basaron en la medicin de los
parmetros siguientes:
a. Volumen de material incorporado
b. Penetracin, d, de la masa al caer dentro de la
columna.

nivel de masa en
golpe precedente
nivel de masa en
golpe actual

Figura 7. Verificacin de la penetracin de la masa.

Por otro lado, se llev a cabo la verificacin de las


columnas terminadas mediante la realizacin de
pruebas de campo realizadas directamente en las
columnas:
Pruebas de presimetro Menard
Pruebas de placa, aplicacin de cargas
superficiales incrementales a travs de placas
metlicas circulares
Las primeras, con el presimetro, se realizaron
para la medicin de los mdulos presiomtricos en
toda la altura de la columna, mientras que las
segundas se realizaron para la medicin de los
asentamientos y la obtencin de las curvas cargaasentamiento-tiempo con las cuales se calcul la
deformabilidad de la columna y se calibraron los
modelos de anlisis.

Figura 8. Pruebas de placa realizadas sobre las CSD.

6 CONCLUSIONES
La Sustitucin Dinmica es una tcnica de
mejoramiento
para
suelos
de
propiedades
mecnicas pobres, que se adeca bien para el
tratamiento de suelos blandos y blandos orgnicos
turba. El estudio de factibilidad de este tipo de
mejoramiento requiere en principio de un anlisis
adecuado y detallado de las caractersticas del
terreno y del sistema de transmisin de cargas para
poder estimar correctamente los asentamientos que
se tendrn en la realidad a corto y largo plazo, as
como de la definicin de las caractersticas de las
columnas a construir: dimetro, profundidad,
separacin, tipo de material de relleno, etc.
En este trabajo se ha presentado la aplicacin
prctica de este tipo de tratamiento en suelos
blandos arcillosos orgnicos depositados sobre roca
caliza en la pennsula de Yucatn, para un proyecto
de construccin de vialidades para la urbanizacin
de un desarrollo habitacional.
Se han descrito las ventajas econmicas que esta
alternativa de este mejoramiento aport al proyecto
contra la solucin estructural original, y se ha
mostrado el procedimiento de construccin realizado
y los controles llevados a cabo en la obra para
aseguramiento del funcionamiento del sistema de
mejoramiento.
REFERENCIAS
Norma NF P 94-160-1. Auscultation d'un lment
de fondation. Parties 1 et 2. Octobre 2000.
Philipponnat Grard, Hubert Bertand. Fondations
et ouvrages en terre. Ed. Eyrolles. Paris, Francia.
2000.
Cassan Maurice. Les essais in situ en mcanique
des sols. Ed. Eyrolles. Janvier 1987.
Magnan Jean-Pierre. Thorie et Pratique des
drains verticaux. Technique et Documentation
Lavoisier. Paris, Francia. 2000.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

Technical Committee

TC-214
3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniera Geotcnica

Noviembre 11-12, 2015 Mxico, D. F.

Transforming marginal land to support a world class development in Panama


Modificacin de suelos marginales para apoyar proyectos de clase mundial en Panam
Gustavo LANGONI1, Roger ARCHABAL1
1Langan

Engineering and Environmental Services

ABSTRACT: The Santa Maria Golf and Country Club development consists of approximately 280 hectares of
low lying coastal wetlands on the western fringe of Panama City in the Republic of Panama. The on-going
development represents one of the most ambitious master-planned developments ever to be undertaken in
Central America. Development of the site presented tremendous geotechnical challenges since the design
flood criteria required that the existing surface grade be raised with approximately 5 meters of fill placed over a
highly compressible marine clay deposit. This equated to nearly 10,000,000 cubic meters of fill to develop the
site. The compressible clay stratum varied in thickness, but was typically between 5 to 10 meters thick. The
required fill loads resulted in over one meter of consolidation settlement. To further complicate the site
development challenge, the developer construction schedule required a site preparation period of less than six
months from fill placement. The theoretical time required for consolidation without ground improvement would
have ranged between 10 and 20 years.
The objectives of this paper are to: 1) present the array of geotechnical design challenges resulting from the
need for thick fills over highly compressible clay; 2) share the geotechnical properties of the site gathered
during the subsurface exploration; 3) present the technical concepts associated with the ground improvement
recommendation consisting of prefabricated vertical wick drains and surcharge to accelerate the consolidation
and precompress the soft clay stratum within the short pre-development period; 4) discuss the use of value
engineering ideas to provide the most economical approaches to implement ground improvement systems, 5)
provide and discuss full scale field settlement readings collected for the pilot test section of the project; 6)
show the value of instrumentation and field measurement in adjusting and improving design recommendations
and in obtaining significant cost savings.

1 PROJECT LOCATION AND DESCRIPTION


The overall site consists of approximately 280 hectares of undeveloped land on the southeast side of
Panama City in the Republic of Panama. Figure 1
shows the location of the project within Panama City.
The overall site is bisected into two parcels, north
and south, by the east-west Corredor Sur highway.
The southern parcel is the subject of this paper and
is approximately 172 hectares in area. It is located in
a low lying coastal area that was originally used for
farming and pasturing activities. Residential and light
commercial development has recently occurred to
the west and north of the site, mangroves and the
Pacific Ocean are south of the site, and similar
undeveloped land is east of the site. The project
components consisted of 1) a championship eighteen

hole golf course, 2) numerous low to high-rise


residential developments, a town center and country
club facilities, 3) artificial lakes, and 4) extensive
infrastructure, including internal vehicular roadways,
underground utility lines, pump stations, perimeter
security and screen walls, etc.
The original grades throughout the parcel were
relatively flat and typically ranged from el +2 m to +3
m MSL, where a soft saturated marine clay material
was typically encountered at the ground surface. The
proposed site grade for the project was on average el
+6.5 m MSL and because of the difference in height
between the original ground surface and the
proposed finished grade; up to approximately 4.5 m
of fill was required to raise the site grade.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

132

Transforming marginal land to support a world class development in Panama

Figure 1. Project Location

These significant areal fills were predicted to


cause an estimated settlement of 1 to 2 m due to the
compressibility and thickness of the soft clay stratum
underlying the project site. Furthermore, these
consolidation settlements were estimated to occur
over the course of more than 20 years after filling in
some cases

Stratum Number

Description

Marine Clay

Intermixed Clay, Silt,


Sand, and Gravel

Weathered to Sound
Shale and Sandstone
Rock Formation

2 SITE SPECIFIC SUBSURFACE CONDITIONS


Thick alluvial soil deposits are present throughout
most coastal regions of Panama. In some areas,
these layers are granular and consist of poorly
graded sand, but in other areas thick layers of marine
clays are found. These finer grained soil deposits
form in depositional areas of rivers and tributaries
creating soft deltaic conditions. Our project site,
which is next to Panama Bay, has thick layers of
normally consolidated marine clays. These clays are
soft and very compressible.
The generalized subsurface conditions throughout
the development generally consisted of the following
inferred strata:

2.1 Marine Clay (Stratum 1)


At the majority of the test boring locations, the native
marine clay was encountered at the existing ground
surface, approximately el +2.5 m. This layer consisted primarily of highly plastic clay with some fractions
of silt and sand in localized areas. The thickness of
this stratum varied considerably from 2 m to 10 m
(averaging about 6 m) and typically trended thicker
from north to south. The stratum was generally very
soft to soft with SPT N-values ranging from weight of
rod (WOR) to less than 5 blows per foot (bpf). Laboratory index property tests were performed for
samples taken from this stratum.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

LANGONI G. et al.

133

* N=SPT N-value (blows/30 cm)

Figure 2. Typical Subsurface Profile

The natural water contents of the stratum varied


typically between 40% and 100%. Liquid limits (LL)
ranged between 30% and 110% (typically over 80%)
and plasticity indices (PI) ranged between 45% and
90% (typically over 55).
Some isolated zones, mostly in the north areas of
the project, had a surficial layer of medium stiff clay,
with SPT N-values ranging from 5 to over 10 bpf.
These were most notably observed along the
northeastern border of the site and near the surface
at some of the borings along the north of the site
adjacent to the Corredor Sur highway. These are
likely isolated high areas that have been hardened by
desiccation, and are not in the area discussed in this
paper.
2.2 Intermixed Clay, Silt, Sand, and Gravel
(Stratum 2)
Below Stratum 1 in some of the borings, primarily on
the eastern portion of the site, a stratum consisting of
intermixed clay, silt, sand, and gravel was encountered. This stratum is typical of highly variable, typically more granular deposits found along the meandering banks of existing or ancient rivers. The
encountered thickness of this stratum varied considerably from approximately 0.5 to 7 m. This heterogeneous stratum varied considerably from very
loose/soft to dense/stiff with STP N-values ranging
from 2 to over 100 bpf.

2.3 Weathered to Sound Sedimentary Rock


Formation (Stratum 3)
Below Strata 1 or 2, a consistent natural rock
formation was encountered at elevations ranging
around el 0 to el -10, but more consistently between
el -5 and el -6.5 in the southern residential
development areas. The sedimentary rock formation
was identified as Shale and Clayey Sandstone. The
rock formation typically has a 1 to 2 m highly
weathered and fractured surface followed by a less
fractured more intact rock formation, based on the
Rock Quality Designation (RQD) values. However,
even the highly weathered and fractured upper rock
zone is competent with the SPT resistance N-values
over 50 bpf to refusal and drill times typically greater
than 15 to 20 minutes per foot. Rock Quality
Designation (RQD) values on the collected rock
cores varied considerably from 0% to 100%.
Typically, the RQD for the upper 5 to 10 m of the rock
is less than 50%. In addition, unconfined
compression tests were performed on selected rock
cores.
Results of the tests show that the
compressive strength of the rock varied from
approximately 72 to 318 kg/cm 2 (1,000 to 4,500 psi).
3 KEY GEOTECHNICAL ISSUES
For the average clay thickness on-site of 6 m and
average fill height of 4.5 meters to reach finished
grade, the magnitude of primary settlement was estimated to be in the range of 1 m. The above settle-

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

134

Transforming Marginal Land to Support a World Class Development in Panama

ments computed were for average conditions, and


varied substantially based on clay thickness. For
these conditions, the time required for primary settlement to occur (without any ground drainage enhancement) would have theoretically been greater
than 20 years which would not be compatible with the
construction schedule and phasing of the development.
Due to the extensive size of the developable area,
the type of soils that had to be improved, and the
need for stable structural support of building
elements as well as surrounding infrastructure, most
ground improvement options were not feasible.
Traditional surcharging alone (fill placed above
finished grade to accelerate settlement), using two to
three meters of fill without wick drains, was
evaluated. For this case, the consolidation time
would be accelerated somewhat to about 6 to 8 years
from the original estimate of about 20 years. For the
proposed residential and roadway developments, the
estimated surcharge time required to achieve primary
settlement would still have been unacceptable and
would not have achieved the scheduling milestones
for the project.
Prefabricated vertical drains (PVDs or wick
drains), installed on relatively close spacings, within
the compressible clay were determined to be the best
option to accelerate consolidation. Wick drains
decrease the drainage path for the pore water within
the clay stratum to dissipate during the consolidation
process. Once the design fill is placed, a surcharge
fill pressure greater than that induced by the
proposed surface improvements, can be placed
above the finished ground surface so as to simulate
the loads of the future construction and preconsolidate or pre-compress the ground. Once
preconsolidated, the ground could theoretically
support structural loads, up to the preconsolidation
stress, using conventional shallow foundations.
4 GROUND IMPROVEMENT DESIGN
The ground improvement system designed for the
site based on the constraints previously outlined
consisted of wick drains and surcharge. The design
details are discussed in the following paragraphs.

Subsequently, we recommended supplemental subsurface exploration be performed and that additional


undisturbed samples be obtained in order to perform
more consolidation tests and better understand the
consolidation parameters.
A total of 15 consolidation tests were performed on
the undisturbed samples collected from several test
borings (identified as LB1, LB1A and LB2). Figure 4
shows the boring location plan with the location of the
consolidation tests.
Nine of the consolidation tests were performed on
traditional horizontally cut samples extracted from the
Shelby tubes. The remaining 6 of the 15 tests were
performed on undisturbed samples cut along the
vertical plane, in an attempt to better understand the
horizontal consolidation behavior and potential
impact on time-rate of settlement. The following
table summarizes the results of the consolidation
tests for the site.
As shown in the following Table 1, the
compression index (Cc) varied from 0.41 to 1.15, with
an average value of about 0.80. The coefficient of
consolidation value (Cv) ranged from 0.38 to 2.54
m2/year with and average weighted value of about
1.39 m2/year. This average value was greater than
the 0.55 m2/year, originally used in our design and
allowed for increased spacing of wick drains based
on the same time constraints.
The horizontal
coefficient of consolidation (Ch) was compared to the
Cv values, but this did not reveal any apparent
behavior trends or relationship. Literature indicates
that Ch is typically greater than Cv, but the samples
tested showed similar coefficient of consolidation
values for both. Hence, we concluded that the
relationship between Cv and Ch within the clay matrix
at the site is relatively close and without any specific
trends.
As stated above, results of the additional
consolidation testing suggested that an improved
coefficient of consolidation (Cv) was appropriate for
the design of the ground improvement system.
Additional engineering analyses was performed to
evaluate the wick drain spacing and surcharge
requirements to meet the project goals, specifically a
preload surcharge period of 6 months using a
coefficient of consolidation (Cv) of 1.39 m2/year. The
analysis yielded a revised wick drain spacing of 1.37
m to achieve a 6 month preload time period.

4.1 Wick Drain Spacing


At the onset of our involvement, our theoretical analysis of wick drains using the Kjellman-Barron formula
suggested a wick spacing of about 1 m in order to
reduce the time for the surcharge fill placement period to about 6 to 8 months (180 to 240 days), for an
average clay thickness of about 6 meters. This was
based on a coefficient of consolidation (Cv) of 0.55
m2/yr, from a previous preliminary geotechnical investigation by others which at that time was the only
consolidation test that was performed at the site.

4.2 Surcharge
Surcharge needs to be placed in order to
appropriately pre-consolidate and pre-stress the
underlying soft clays to stress levels above those
imposed by the future surface improvements
(primarily residential structures and roadways). The
surcharge fill required to pre-compress the clay to
anticipated post-construction stresses was generally
2 m high (over proposed finished grade) in the
residential development areas and 1 m high (over

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

LANGONI G. et al.

proposed finished grade) in the roadway areas.


Appropriately performed, the surcharged ground
would allow for normal construction of the future

135

residential structures using conventional shallow


foundations and typical construction techniques.

Figure 3. Boring Location Plan.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

136

Transforming Marginal Land to Support a World Class Development in Panama

Table 1. Summary of Laboratory Consolidation Tests.


Boring
ID

Sample

Description of M aterial

Depth (m)

Water
Content
(%)

LB-2

ST-1A

CH, dark brown CLAY, trace f. sand

0.6

42.6

89

27

LB-2

ST-1A

CH, dark brown CLAY, trace f. sand

0.6

40.5

89

LB-1

ST-1B

CH, gray CLAY, trace f. sand

1.1

48.8

95

LB-2

ST-2B

CH, dark brown CLAY, trace m -f sand

1.7

LB-2

ST-2B

CH, dark brown CLAY, trace c-f sand

ST-3C

CH, gray CLAY, trace m -f sand; clay


nodules noted

LB-1A

ST-3C

CH, dark brown CLAY, trace m -f sand; clay


nodules noted

LB-1A

ST-4B

LB-2

LL PL Plasticity
(%) (%) Index (PI)

Cv
(m2 / yr)

Cc

Cr

62

0.317

0.083

0.001

0.38

27

62

0.300

0.062

0.001

0.44

29

66

0.413

0.090

0.003

0.38

89.1

111 33

78

0.833

0.071

0.008

0.89

1.7

88.1

111 33

78

0.948

0.138

0.009

1.19

2.4

88.9

116 26

90

0.826

0.080

0.008

0.45

2.4

88.6

116 26

90

0.827

0.070

0.008

0.86

CH, gray CLAY

2.8

95.3

100 29

71

1.007

0.106

0.015

0.51

ST-3B

CH, gray CLAY; sand pockets and shell


fragm ents noted

3.0

75.6

77

31

46

0.636

0.046

0.009

1.79

LB-2

ST-3B

CH, gray silty CLAY; som e f. sand;


num erous sand layers and pockets noted

3.0

45.0

n/a

n/a

n/a

0.378

0.011

0.002 190.67

LB-2

ST-4A

CH, gray CLAY, f. sand; silt lenses noted

4.0

76.2

88

29

59

0.662

0.073

0.009

1.34

LB-1A

ST-5B

CH-OH, gray ORGANIC CLAY, trace f.


sand; silt pockets and shell fragm ents
noted

4.6

84.5

87

30

57

1.042

0.115

0.011

2.54

LB-1A

ST-5B

CH, gray CLAY, trace f. sand; organic


m aterial noted

4.6

96.5

87

30

57

1.148

0.109

0.014

1.54

LB-1

ST-2B

CH, gray CLAY; trace f. sand; shell


fragm ents noted

4.8

81.9

85

31

54

0.708

0.086

0.008

0.64

ST-2B

CH, gray CLAY; trace f. sand; silt pockets


noted

4.8

88.8

85

31

54

0.721

0.090

0.008

0.63

LB-1A

LB-1

5 FULL SCALE PILOT TEST PROGRAM


A full-scale pilot test program was constructed and
instrumented at the site to better understand the
time-rate consolidation behavior of the underlying
compressible strata under varied wick drain spacing
and surcharges.
5.1 Test Section
Construction of the pilot test area began on 18 April
2008 with the installation of the first wick drains. The
pilot test area, shown in Figure 4, was located in a
representative area of the project. Test boring L10E
was performed within the pilot test area to verify the

thickness of the clay prior to the test program, and


indeed it verified a representative clay thickness of 6
m.
The test area was divided into six zones to verify
the settlement changes from three different wick
drain spacings of 1.00 m, 1.37 m, and 1.75 m and
the two surcharge heights specified for the project of
1 m and 2 m, for roadways and residential areas
respectively.
Two settlement plate monitoring
devices were installed in each of the varied
surcharge and varied wick spacing sections. Hence,
a total of twelve settlement plates were installed. In
addition, six piezometers were also installed to
measure the changes in pore water pressures with
time.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

LANGONI G. et al.

137

Figure 4. Full Scale Pilot Test Section.

Monitoring of settlement plates installed within the


pilot test area was semi-weekly beginning in April
2008 and continuing until February 2010. Figure 6
graphically presents the settlement data on a linear
time scale. The graphs show the varied wick
spacings (1.37 m, 1.50 m, and 1.75 m) within the 1
meter and 2 meter surcharge zones, respectively.

Figure 7 presents the settlement data on a


logarithmic time scale in order to better evaluate
when primary consolidation is transitioning into
secondary compression or creep.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

138

Transforming Marginal Land to Support a World Class Development in Panama

Figure 5. Settlement Data on a Linear Time Scale.

Figure 6. Settlement Data on a Logarithmic Time Scale.

Subsequent to surcharge top-out, settlements of


approximately 90 cm and 100 cm within the 1 m and
2 m surcharge zones, respectively, were recorded.
Readings after 160 to 180 days showed a
significantly reduced settlement trend. The point at
which the reduced settlement trend begins is
characteristic of the completion of primary

consolidation and on-set of the long-term secondary


compression phase. The pilot test settlement data
obtained closely corroborated the original theoretical
estimated settlement values of 90 to 100 cm.
Based on the settlement data obtained from the
test section, the transition time from primary
consolidation to secondary compression (after

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

LANGONI G. et al.

surcharge top-out) proved to be close to the original


estimate of 180 days using a wick drain spacing of
1.37 m. The similar settlement time frame that
occurred for the wick drain spacings tested, however,
suggested that the drainage properties were affected
by the scale of the large test area. Although the
cause for this has not been clearly understood it is
likely attributed to drainage paths within the soil and
wick drains due to the heterogeneous nature of the
soil mass under the full scale section. This prompted
the construction of a second test area and an
adjustment of the wick drain spacing to 1.75 m and 2
m in some areas of the project saving millions of
dollars in ground improvement costs.
6 CONCLUSIONS
Ground improvement using wick drains and
surcharging was successful in pre-consolidating the
highly compressible ground in a short time frame in
order to allow for construction in about 180 days post
surcharge top-out. A full scale field test section
proved to be a cost-effective and valuable tool to
allow for optimization of wick drain spacing. Wick
drain spacing and improvement times calculated by
theoretical methods using consolidation data and test
borings appear to err on the conservative side. This
is likely because the theoretical methods do not
account for the effect of thin sandy soil seams within
the compressible clay layer that have a higher
hydraulic conductivity and enhance the global
drainage capability. Full scale field test sections can
capture the actual drainage properties of an entire
area in a way that cannot be evaluated by a

139

subsurface exploration program. Properly performed,


field test sections represent a very minor
instrumentation and evaluation cost with a very large
economic benefit in reduced ground improvement
costs.
For the Phase 1 construction, surcharging to
above the design loads for the project allowed for
successful construction of both the significant site
infrastructure as well as a significant number of low
rise residential structures on conventional shallow
foundations. Long term creep settlements were
measured to be on the order of 1 cm for the first year
of monitoring. Although the monitoring period was
not sufficient to make a definitive statement about
long term creep settlement, the readings after the
first year showed little signs of creep movement and
are expected to decrease with time. Further, the
settlements measured by monitoring pins on actual
constructed structures were negligible and consistent
with results from the full scale test sections.
REFERENCES
Dhar, A.S., Ameen, S.F. (2001). Ground
Improvement Using Pre-loading with Prefabricated
Vertical Drains.
Holtz R. and Kovacs W. (1981). An Introduction to
Geotechnical Engineering.
Lee et. al. (2001). Soft Soil Engineering, p.4-8.
Terzaghi K., Peck R., and Mesri G. (1996). Soil
Mechanics in Engineering Practice, Third Edition, p.
100-116.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

3rd International Conference on Deep Foundations


Deep Foundations and Soil Improvement in Soft Soils

Technical Committee

TC-214

Author Index
Page

Page
ARCHABAL Roger
ARENAS Fernando
AUVINET-GUICHARD Gabriel
CHATTE Rmi
CHIARABELLI Marco
CIRION ARANA Alfredo
CUEVAS Alberto
DEMING Peter W.
GERRESSEN F.
HUANG Yanbo
IBARRA Enrique
LANGONI Gustavo
LIN Cheng
LIN Guoming
LPEZ Germn
MARINUCCI Antonio
MASSOUDI Nasser
MENDOZA Manuel J.
MOTHERSILLE Devon
NIKOLAOU Sissy
OKUMUSOGLU Bora
OROZCO Marcos
PAGLIACCI Federico
PANIAGUA Walter
PAULN AGUIRRE Juan
POLETTO Raymond J.
PRADEL Daniel

131
85
51
127
19
127
85
39
117
3
77
131
3
3
31
19
93
77
97
39
97
77
67
31
127
39
107

RODRGUEZ-REBOLLEDO
Juan-Flix
RUFIAR Miguel
SEGOVIA Jos
SLIWOSKI Richard
TAKUMA Takefumi
TAMARO George J.
TIWARI Binod
WILK Charles M.

SOCIEDAD MEXICANA DE INGENIERA GEOTCNICA A.C.

51
77
31
93
11
39
107
123