Ground

Improvement
Third Edition

Edited by Klaus Kirsch and Alan Bell

Boca Raton London New York

CRC Press is an imprint of the

Taylor & Francis Group, an informa business
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2013 by Klaus Kirsch and Alan Bell
CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Version Date: 20121210

International Standard Book Number-13: 978-0-203-83897-6 (eBook - PDF)

This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information stor-
age or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copy-
right.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222
Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that pro-
vides licenses and registration for a variety of users. For organizations that have been granted a pho-
tocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Preface

Ground improvement techniques continue to progress in addressing ground

engineering problems across the world, particularly in urban areas where
land development and reuse need to be eficient not only in the geotechni-
cal engineering but in time, cost, and energy used. As well as in expanding
markets, recent growth has also been seen across a range of methods, in
increasing productivity due to investment in plant and equipment, and in
improvement in technical performance and quality due to electronic moni-
toring and control methods. Ground improvement methods are also fre-
quently able to demonstrate low carbon impact and excellent sustainability
credentials as these issues become more important.
The third edition of this well-known book provides a comprehensive
overview of the major ground improvement techniques in use worldwide
today. The chapters are fully updated with recent developments and have
been written by recognised experts who bring a wealth of knowledge and
experience to bear on their contributions.
Ground Improvement is written for civil and geotechnical engineers
and for contractors involved in piling and ground engineering of any kind.
Advanced graduate and postgraduate civil engineering and geotechnical
students will ind the book most helpful in guiding their studies.

v
Contents

Preface v
Editors vii
Contributors ix

1 Introduction and background 1

ALAN BELL AND KLAUS KIRSCH

2 Deep vibro techniques 17

JIMMY WEHR AND WOLFGANG SONDERMANN

3 Dynamic compaction 57
BARRY SLOCOMBE

4 Prefabricated vertical drains 87

JIAN CHU AND VENU RAJU

5 Permeation grouting 169

GERT STADLER AND HARALD KRENN

6 Jet grouting 207

GEORGE BURKE AND HIROSHI YOSHIDA

7 Soilfracture grouting 259

EDUARD FALK AND CLEMENS KUMMERER

8 Compaction grouting 299

JAMES HUSSIN

iii
iv Contents

9 In-situ soil mixing 329

MICHAŁ TOPOLNICKI

10 Dry soil mixing 435

MARCUS DAHLSTRÖM
Chapter 1

Introduction and background

Alan Bell and Klaus Kirsch

CONTENTS

1.1 Purpose of ground improvement processes ....................................... 1

1.2 What the book covers....................................................................... 2
1.3 Historical development..................................................................... 4
1.4 Health, safety, and environmental considerations ............................ 6
1.4.1 Site mobilization and demobilization..................................... 7
1.4.2 Hazardous materials.............................................................. 8
1.5 Greenhouse gas emissions................................................................. 8
1.6 Compaction by blasting.................................................................. 10
1.7 Ground freezing.............................................................................. 12
References ............................................................................................... 14

1.1 PURPOSE OF GROUND

IMPROVEMENT PROCESSES

When faced with dificult ground conditions at a project site, an engineer has a
number of possible strategies to employ in order to achieve the project objectives.
The most obvious is to ind another site, but this is only very rarely practicable.
Pressure on land, the need to use poor sites, and the location of many cities
in estuaries or river situations make this option increasingly dificult. Another
option is to redesign the building or structure to accommodate the prevailing
dificulties arising from the ground, and where possible this is a good solution.
Yet another possibility is to remove the troublesome ground and to replace it
with more suitable material, and this can often be cost effective providing the
depth to be addressed and the quantities concerned are relatively small.
If none of these avoiding strategies are technically or economically real-
istic, then the prevailing ground conditions must be addressed. A common
potential solution is to adopt a system such as piling, in order to bypass
the dificult ground and found in suitable material. However, this can be
expensive and time consuming and may actually be dificult to achieve in

1
2 Ground improvement

very deep ground. In addition, for some classes of geotechnical problem

such as tunnelling, piling may be unsuitable. For such reasons improving
the ground to achieve an appropriate engineering performance is an increas-
ingly successful approach worldwide when faced with problem ground con-
ditions, partially evidenced by the two earlier editions of the present book.
Ground improvement is normally understood as the modiication of the
existing physical properties of the ground beneath a site to suficient depth
to enable effective, economic, and safe permanent or temporary construc-
tion in practical timescales. Typical objectives would be one or a combina-
tion of the following:

(1) An increase in shear strength or density to improve bearing capacity

or to provide suficient support for excavations or tunnels
(2) A reduction in compressibility to minimise total or differential settle-
ments of buildings or structures, or other deformations in the ground
arising from excavation or tunnelling
(3) A reduction in permeability to minimise low of ground water to pre-
vent inundation or water damage or to isolate zones of contaminated
ground water
(4) Conversely, an improvement in deep drainage in order to assist pre-
loading or surcharge techniques
(5) Controlled displacement of the ground in order to dispel previous
differential settlements or ground distortions, or to compensate for
ground movements arising from excavation or tunnelling
(6) Prevention of liquefaction or reduction in lateral spreading beneath
or near both new and existing structures during earthquakes, emp-
loying densiication, replacement with stronger materials, or deep
drainage

The ground improvement processes used to deliver these objectives form

the subject matter of this book, as set out in Section 1.2.
It should be noted that recent environmental legislation coupled with the
need to recycle previously developed sites has led to considerable growth in a
different concept of improving the ground—namely, to minimise or remove
the hazardous effects of sites contaminated by toxic waste or chemical by-
products from industrial processes. This subject is beyond the scope of this
book and the reader should refer to the specialist literature on these topics.

1.2 WHAT THE BOOK COVERS

This third edition of Ground Improvement will provide the reader with a
sound basis for understanding and further study of the most widely used
processes for ground improvement.
 Introduction and background 3

Developments in equipment and methods have continued apace since

the publication of the second edition, and where relevant are included in
the ensuing chapters of this book. Indeed, the editors are grateful to the
authors of the following chapters in the book, all of whom are recognised
experts in their respective ields. Their contributions provide an overview
of the processes concerned and the key geotechnical and design consider-
ations involved, together with details of the equipment needed for success-
ful execution. The methods are well illustrated with relevant case histories
revealing applications in practice.
Since soil strength and compressibility are highly inluenced by the par-
ticle packing or density in most engineering soils, densiication is a useful
approach. In granular soils this is frequently achieved most eficiently using
vibratory methods to force particles into more closely packed conigurations.
Methods employing tools in which the vibrator can be taken deep into
the ground are very eficient and are described in Chapter 2. Another
important global technique is dynamic compaction, which employs large
weights dropped from height to create the compactive energy needed, and
is dealt with in Chapter 3.
In cohesive soils, an increase in shear strength and reduced compress-
ibility can be achieved by consolidation, usually achieved by direct loading.
The process is time dependent and can be hastened and better controlled
using deep prefabricated drains and associated methods. These are compre-
hensively dealt with in Chapter 4.
The remaining six chapters in the book deal with various techniques
involving the injection of materials into the ground in order to provide geo-
technical improvement of various kinds. Chapter 5 describes permeation
grouting, which involves the displacement of the ground water or air in
soil pores or rock issures, using an agent, usually termed grout, which is
suficiently luid to permeate the ground. This agent subsequently hardens
to create the intended improvements. Jet grouting is covered in Chapter 6.
This technique uses powerful jets of grout, or grout with other luids to dis-
place or mix with the ground. In this way zones of ground with increased
strength or stiffness, or barriers to low can be formed. Soil fracture grout-
ing is a displacement technique and employs inely controlled injection of
relatively thin but multiple veins of grout to address excess building settle-
ment, to lift structures, or to compensate for ground movement during tun-
nelling. It is described in Chapter 7. Compaction grouting, developed in the
United States and now used around the world, is dealt with in Chapter 8.
This is also a displacement method and can be used to compact or rein-
force the ground with introduced grout. Recent changes in terminology
originating in the United States are helpfully explained in detail. In-situ soil
mixing processes continue to develop worldwide, and Chapter 9 provides
comprehensive coverage of the main deep mixing processes in use across
the world and describes several new techniques, such as the trench-mixing
4 Ground improvement

TRD and the panel-mixing CSM approaches. Several new case histories are
also included. Chapter 10 covers dry soil mixing using the Scandinavian
approach, as this process has continued to see worldwide application.
In the remainder of the present chapter several topics which would not
justify a separate chapter are included, such as the history of the two main
means of creating improvement (see Section 1.3). Brief notes on health and
safety for ground improvement sites are included in view of its common
relevance (Section 1.4). The effects of ground improvement on greenhouse
gas emissions have increasingly been addressed since the second edition and
now warrant inclusion (Section 1.5). Overviews of two ground improve-
ment techniques which, by their nature, have somewhat limited ranges of
application have also been included. Blasting can be an effective means of
improving granular soils by densiication and is described in Section 1.6.
The only reversible ground improvement process of ground freezing can be
powerful where applicable, and is outlined in Section 1.7.

1.3 HISTORICAL DEVELOPMENT

Since earliest times humankind has found ways of dealing with poor
ground in order to form pathways and later roads using such simple strate-
gies as placing beds of reeds or saplings to support the weight of people
and animals over soft ground. It is only relatively recently that the means
of engineering dificult ground by compaction, consolidation, or by adding
materials by permeation or mixing has seen signiicant advances. These
processes developed during and after the Industrial Revolution, but mainly
in the early twentieth century. In these years, better understanding of soil
mechanical behaviour emerged through the work of Terzaghi and others;
practical ground investigation became possible; and equipment and materi-
als development reached the stage that signiicant volumes of soil could be
treated. Two main approaches are in worldwide use today, namely deep
vibratory treatment and injection or mixing of grouts. A brief historical
review of these two topics follows in view of their importance.
Kirsch and Kirsch (2010) describe how depth vibrators were developed
in Germany in the 1930s. Initially aimed at concrete densiication, they
soon were applied more effectively for sand compaction. Consequently,
deep vibratory stabilisation for both natural and illed cohesionless soils
was used widely for a range of applications in Germany and further aield,
particularly after 1945. A further major area of application was added in
1956 when depth vibrators were employed to form stone columns in silty
materials, leading to the application to cohesive soils more generally.
Further development and improvement of the special plant and equip-
ment necessary for the execution of this ground improvement method
together with the experience gained in practice considerably increased the
 Introduction and background 5

range of application in foundation engineering, notably after the advan-

tages of the method were recognised mitigating the liquefaction potential of
soils in earthquake-prone regions by densiication and/or drainage. Today
the method can effectively be designed for its purpose and can be well con-
trolled during its execution. It is interesting to note that irst steps towards
eventual process automation are already being used in practice.
The idea of injecting cement slurries, known as grouts, into the ground
to improve their engineering characteristics also saw early development. It
is believed that grouting of subsoil was irst performed more than 200 years
ago in France (about 1810) by the French engineer Charles Bérigny, using
a suspension of pozzuolana cement in water to stabilise alluvial deposits
forming the foundations of a bridge (Glossop 1961). Further development
of the grouting process (procédé d’injection) in the nineteenth century was
by the introduction of new hydraulic binders, particularly the invention of
Portland cement in 1821. The method was already well developed at the
outset of World War I with the use of pumps, pressure control, and the need
for iltration all established.
Direct injection of simple cement grouts into the ground was often ham-
pered by failure to permeate the ground, either because the pore size distri-
bution of the soils were small in comparison to the grading of the cements
or because the methods of injection, often from open-ended casings, were
too crude. Low pressure permeation grouting using simple cement grouts is
usually limited to gravel containing perhaps some coarse sand.
Important steps forward in addressing this limitation were accomplished
by attention to the materials used for injection. Dutchman Hugo Joosten
in Germany, by his invention of the Joosten system, used chemicals in the
form of highly concentrated sodium silicates and calcium chloride as grout
material to form precipitate silicates in situ to treat sandy cohesionless soils
(Joosten 1926). These much iner-grained grouts could permeate more
readily than simple cements. The method was then widely used in Berlin in
the construction of the underground railway.
By the late 1950s a single shot approach was developed by mixing organic
hardeners and sodium silicate before injection. Today various proprietary
re-agents are available as hardeners, with widely differing properties.
Another development involved creating, by ine grinding, so-called micro-
ine or ultraine cements, which also allow permeation into coarse sands
or sometimes even iner soils, and these are also in use today. Various new
chemical formulations were also developed in the 1960s and 1970s which
enabled even iner grained cohesionless soils to be treated, but with limited
application today due to concerns about toxicity.
Developments for dealing with some of the limitations of soil grouting
also came from improvement in equipment. The invention of the tube à
manchette pipe or TaM pipe (Ischy 1933), still very much employed in
grouting processes today, was very signiicant. These pipes, consisting of
6 Ground improvement

grout ports with rubber sleeve valves, are placed in boreholes to attain the
depths required, and are grouted in place using a relatively weak sleeve
grout. One set of ports can be isolated at a time, and grout can be injected
into the surrounding ground after it expands the surrounding seal and
breaks the sleeve grout. This enables control of grout volume or pressure
during injection at speciic points in the subsoil. Littlejohn (1993) provides
a useful summary of the history of injection processes.
Subsequent grouting development has concentrated more on the develop-
ment of entirely new ground improvement processes using simple cement
grouts, partly due to concerns over toxicity of some chemical grouts and
partly to the desire for improved performance. By the early 1990s jet grout-
ing; compaction grouting; soilfracture grouting; and soil mixing methods
were all widely and successfully applied in addition to permeation grouting
(e.g., Bell 1994). Since then there has been increased use of all of these, and
soil mixing in particular is now more widely used. Further technical devel-
opment of all grouting methods has continued, particularly in relation to
electronic monitoring and control on site. Some history on these methods is
included in Chapters 6 through 9.

1.4 HEALTH, SAFETY, AND ENVIRONMENTAL

CONSIDERATIONS

In recent years the construction industry worldwide has seen signii-

cant improvements in the safety of construction workers and the public.
Legislation, formal management systems, and motivational training have all
played their part. Indeed, safety is a critical component of all construction
in general and ground improvement in particular as it inevitably contains
aspects that are potentially unsafe. The specialist piling and ground improve-
ment industries are very committed to safety and minimising the environ-
mental effects of construction. For example, the European Federation for
Foundation Contractors (EFFC) holds the attainment of the highest stan-
dard of safety as a key objective and its members have established a health
and safety charter, together with publications and advice on the subject.
Extensive procedures involving risk assessments, work instructions,
method statements, and training, which together form site-speciic project
safety plans, are now commonly used in minimising safety and environ-
mental hazards through general and specialist ground improvement con-
tractors, as well as client and public bodies. The following provides some
limited comment on the health, safety and environmental impact of ground
improvement processes as a brief introduction to the subject in view of its
consistent importance. However it cannot be a comprehensive presentation
of the subject, and reference needs to be made to local regulations and writ-
ten procedures, and the specialists in the particular processes.
 Introduction and background 7

1.4.1 Site mobilization and demobilization

In common with all site operations, health and safety considerations form
an integral part of establishing construction activity and leaving after proj-
ect completion. Clear delineation is needed for site entrances and exits,
temporary roads for materials’ supply trucks, pedestrian walkways, site
boundaries with protection and exclusion of the general public, and clear
storage and load/unload areas. The working surface, suitably lit, for all
plant and equipment should be engineer-designed and capable of maintain-
ing support in all weather. Simple site procedures can be used to ensure this
work is done to the appropriate standard prior to commencement (e.g., by
the use of a working platform certiicate). If the platform is not integrated
into the inal works, a plan for its disposal is required. Overhead power
lines and underground services need to be clearly identiied and delineated,
together with instructions as to avoidance or minimum clearances given by
appropriate authorities prior to commencing work. Measures need to be
introduced so that noise and dust are minimised and kept within agreed
limits for the general public beyond the site boundaries, and for construc-
tion personnel on site.
Site operatives are required to be trained and experienced with certi-
ied skills, or if in training have adequate supervision from suitably experi-
enced colleagues. Safety equipment supplied must be worn at all times and
employed in accordance with training and advice given. Often the processes
require physically lifting materials such as cement bags, or other heavy
objects such as hoses or steel casing. Speciic training in proper lifting pro-
cedures and speciied max loads for any lift should be taken as a minimum
approach. It is important to employ lifting devices such as winches and
crane arms in all cases where limits are exceeded, and these may in some
cases form part of the drilling equipment
Operational hazards on site must be identiied in advance and plans put
in place to deal with them. Some ground-improvement processes generate
spoil from the ground and this needs to be controlled to prevent injury near
drilling or boring equipment, or from lying debris necessarily generated on
dynamic compaction sites. The spoil must be controlled on site to minimise
deterioration of working and access areas, and measures taken for its safe
re-use or disposal so as not to contaminate the immediate environment.
Trip hazards such as open boreholes must be clearly marked on site.
The plant and equipment used to perform any ground improvement pro-
cess needs to be in good operating condition. Since many pieces of equip-
ment are often used together to enable the process to be eficiently executed
on site, proper consideration needs to be given to all items including any
attachments, not only the large plant. Safety precautions and operating
instructions are generally published for each piece of equipment, includ-
ing initial preparation on site, and should be reviewed prior to use and
8 Ground improvement

updated in the event of modiications or the introduction of revised items

of equipment. Regular maintenance, inspection, and certiication of all
equipment at agreed intervals are needed to ensure continued safe operat-
ing equipment.
Measures including automatic cutoff devices or guarding should be in
place to prevent injury to operatives near to rotating drilling or boring
equipment. The correct mode for moving and operating equipment must be
made clear, and in moving there must be clear guidance to ensure the safety
of adjacent personnel.
One aspect particular to grouting processes is the condition of the
grout hoses and hose connections. It is important that these be rated to
safely withstand the pumping pressure, be operated properly and regularly
checked and certiied. Failure of the hoses or hose connections can result
in the high-energy release of grout, potentially resulting in severe injury.
Also, whipping high energy hoses for grout, air, or other luids are highly
dangerous, and whip checks at connections should be used. For both air
and pressure grout equipment, clear procedures for pressure release in any
circumstances including cleaning, must be identiied and adhered to.

1.4.2 Hazardous materials

Material safety data sheets for all materials to be used on the project should
be reviewed and training given prior to beginning work. Cements and other
cementitious materials or chemicals are commonly used in grouting and
soil mixing and other processes and are very caustic. Prolonged exposure
to the skin or eyes can cause severe chemical burns and permanent injury.
Consequently, risk assessments and the use and enforcement of safe work-
ing methods are vital. Hazards inherent from the design approach are also
to be considered in minimising injury or illness in site personnel and the
general public.

1.5 GREENHOUSE GAS EMISSIONS

In 1997 the Kyoto protocol was ratiied by participating nations, and is

a treaty aimed at stabilising greenhouse gas concentrations in the Earth’s
atmosphere at a level to limit anthropogenic (human) interference with the
climate system. Indeed, there is now worldwide awareness, and acceptance
in the scientiic community, of the greenhouse gas effect on global climate.
The construction industry, in common with other industries, is conse-
quently looking at its own emissions, so that these can be better under-
stood and consequently minimised or even eliminated. Methodology and
databases are now available for identifying and calculating the key inputs.
This is often done by employing the concept of carbon dioxide equivalent to
 Introduction and background 9

the key greenhouse gases released by the process (for example, Hammond
and Jones 2011).
Calculations can also be used to compare different ground improve-
ment and other geotechnical processes such as piling methods, and this
can be important in minimising the total carbon dioxide equivalent for a
given project. Wintzingerode et al. (2011) list seven key potential sources of
greenhouse gas emissions for ground improvement systems, namely:

r Raw materials
r Transport for materials
r Transport for personnel
r Transport for equipment
r Product manufacture
r Waste or spoil emissions
r Transport for waste or spoil

Such calculations can be used to examine different components of the

construction process, and these clearly show the large impact of manu-
factured construction materials such as Portland cement and steel, with
other inputs, notably the energy requirement for construction plant and
equipment, usually much lower than for materials, as seen in comparative
databases (e.g., EA 2010, GEMIS 2010). The other emissions are often very
small for ground improvement projects. Nevertheless, each project must be
studied separately. For example, the degradation of spoil consisting of peats
or highly organic soils removed from below the water table can generate
relatively large emissions (Hall 2006).
Several bases for comparison are possible and it is important to understand
the implications of these. For example, Zöhrer et al. (2010) compare different
methods using MJ/lineal metre for comparison. This provides a means of com-
paring a wide range of products, and is illustrated in Figure 2.18 in Chapter 2.
The strong conclusion was the very low impact of the vibro processes by com-
parison with the others due to the use of quarried materials with this system.
Egan and Slocombe (2010) used several actual foundation projects for
their basis of comparison and found that replacement or partial replacement
of piles with vibro stone column ground improvement systems resulted in
between 92.5%–96% reduction in embodied carbon dioxide, and where
piles could only partially be replaced a reduction of 36.4% was noted.
Wintzingerode et al. (2011) draw attention to comparing the total emis-
sions per unit load carried, and also the total emissions per square metre
of inal construction. Examples illustrate the large reductions of about a
factor of 11 in carbon dioxide equivalents gained by switching from bored
piles to vibro stone columns.
Pinske (2011) compared ive different ground improvement methods
(deep soil mixing, vibro replacement, vibro compaction, deep dynamic
10 Ground improvement

compaction, and earthquake drains). All were compared on the basis of

a functional unit of treating 25,000 cubic meters (50 m × 50 m × 10 m)
of loose, sandy hydraulic ill, intended for use at a speciic site. Deep soil
mixing was the most impactful method, mainly due to the use of Portland
cement. However, introduction of slag cement reduced greenhouse gas
emissions to nearly a quarter of the conventional method. A combination
of vibro replacement stone columns, deep dynamic compaction, and vibro
compaction resulted in the lowest environmental impact based on green-
house gas emissions and life cycle energy, as with the previous examples
primarily because these methods do not employ manufactured materials.
These studies indicate that ground improvement methods generally can
offer environmental beneits as well as technical and cost advantages over
other approaches. Systems such as vibro compaction, vibro stone columns,
and dynamic compaction, which do not employ manufactured materi-
als, have a clear advantage in many situations. Deep drains, soilfracture
grouting, blasting, and freezing use relatively low amounts of manufac-
tured materials and often develop relatively low emissions. Permeation
grouting, jet grouting, compaction grouting, and deep soil mixing all
use larger quantities of manufactured materials relative to other ground
improvement methods, but can still show lower impacts than other geo-
technical approaches such as piling, depending on the project. In addition,
low impact cements, slag mixes, and other low impact materials are being
developed and increasingly used to further enhance the sustainability of
such methods.
However, studies also show that each site must be considered on its mer-
its, with the technical requirements and the prevailing ground conditions
primarily determining which systems are appropriate, at which stage the
most sustainable from among these can be identiied.

1.6 COMPACTION BY BLASTING

Densiication of granular soils can be achieved by detonating explosives

in the ground. Following ield tests carried out in the former USSR in the
1930s, deep compaction by blasting became known as a method of ground
improvement through publications of Abelev and Askalonov (1957) and
Ivanov (1967). The treatment of loose granular soils by blasting is based
upon a sudden dynamic load stimulating the grain structure to rearrange,
reduce its porosity, and ind a closer density. Not unlike the behaviour of
granular soils when subjected to dynamic forces, deriving from the impact
of heavy weights when dynamic compaction is carried out or during earth-
quakes, the prerequisite for eficient compaction is full water saturation.
To date the method of compacting loose granular soils below the water
table by explosives has been used worldwide with positive results. It can be
 Introduction and background 11

executed at relatively low costs generally for very large volumes of granular
deposits. However, it does not yield such high densities as those achieved
by vibro compaction with its scope of application being a moderate but
rather homogeneous densiication (Smoltczyk 1983). It should be noted that
directly at the location of the charge, some heterogeneity or loosening can
arise, particularly if the ines content in the soil is high.
The method can be used from irregular surface conditions and from
ground unsuitable for heavy plant, but it is restricted to loose granular soils
with low silt contents. Only occasionally the method has also been used to
compact ine grained soils, such as loess. The depth range of the method is
generally beyond the reach of vibro compaction, as has been reported by
Solymar et al. (1984). The effective range of the explosion impact deter-
mines the placement of the charges in the ground. It is between 10 m for
low charges of 10–15 kg of TNT equivalent per bore hole and up to 20
m for charges of 30 kg. Placement of the charges is generally in boreholes
often supported by bentonite slurry.
Similar to vibro compaction and the effect of an earthquake, the explo-
sive impact results in shear waves leading to partial or even total liquefac-
tion of the granular deposit and ultimately to a densiication of the soil
when the pore water overpressure has dissipated. The density that can be
achieved by the blasting method depends of course on the explosive energy,
the distance of the bore holes containing the charges in the ground, and on
the same soil characteristics determining the suitability of granular soils
for vibro compaction. The improvement expressed by the increase in rela-
tive density Dr is generally ΔDr = 15%–30%, in exceptional cases more,
particularly when the original density is very low. The extension of the
compaction reaches 20%–50% deeper than the installation depth of the
explosive charge. The success of deep compaction by blasting also depends
on the layout of the blast holes and the sequence of the ignitions in multiple
blasting. Quality control measures are similar to those applicable for vibro
compaction. In the absence of a reliable theory and simple design rules the
method application relies on experience and on trial compactions ahead of
any contract work. Gohl et al. (2000) have produced a promising theory
based on cavity expansion theory and compare with experience on nine
projects.
Densiication arising from compaction blasting of loose water-saturated
granular soils is strongest, whereas the effect on denser deposits tends to
be less. This behaviour leads to an equalisation of density and homogenises
the granular deposit. Although the method is regarded as an economical
means of compaction, its application remains scarce, probably because the
use of explosives for subsurface blasting requires special permissions that
are not easy to obtain. In addition, the environmental impact is substantial:
the emission of noise and far-reaching shock waves needs to be investigated
and controlled throughout the project, as does the emission of gases and
12 Ground improvement

fumes from the explosion, which are injected into the ground and may con-
taminate the ground water.
The need to rehabilitate large deposits in the open brown coal mining
areas in eastern Germany has resulted in an extensive use of the compac-
tion blasting method in parallel to vibro compaction. It has triggered inter-
esting ield trials of blasting and intensive research work on this method
of ground improvement (Kolymbas 1992, Raju 1994, Tamaskovics 2000).
However, its application has been diminishing at this site due to safety con-
siderations. General descriptions of the deep blasting ground improvement
method can be found in Damitio (1970), Mitchell (1981), Kolymbas (1992),
Gohl et al. (2000) and Gambin (2004).

1.7 GROUND FREEZING

The artiicial freezing of soil has been known for well over 100 years.
The method was patented in 1883 by Poetsch in Germany but previously,
in 1862, practiced in the UK (Harris 1995, Jessberger and Jagow-Klaff
2003). By this method water-bearing soils are chilled to such an extent
that the pore water freezes, providing the frozen soil with considerably
higher strength than in its original state and rendering it at the same time
impermeable to water. These changed soil conditions are transient and
reversible as the soil returns to its original characteristics when thawing,
provided that no changes in water content occur, which may happen in
cohesive soils due to the development of ice lenses. Soil freezing is therefore
the only reversible method of ground improvement, and it requires a con-
tinuous supply of energy during its application to maintain the necessary
soil temperature and desired state (for example, its stiffness, strength, or
impermeability).
The method was originally developed to sink large mining shafts through
water-bearing soils. Only relatively recently soil freezing has also been used
in tunnel construction and to resolve dificult problems in ground engineer-
ing, often as the method of last resort (closure of leakages in water barriers;
retrieving of artefacts or valuable machinery from dificult ground condi-
tions). It is also used in soil investigation measures to obtain undisturbed
samples of saturated noncohesive soils to measure their density. Very
recently the method has also been proposed in environmental engineering
to freeze (encapsulate and immobilise) and subsequently safely remove haz-
ardous soil. Its application is generally restricted in time.
Soil freezing is achieved by taking heat away from the ground, generally
using either of the two methods:

r Freezing by brine (CaCl 2) circulation with refrigerators

r Freezing with liquid nitrogen (LN2)
 Introduction and background 13

Both methods require special piping to be installed in the soil for the
introduction of the coolant. Liquid nitrogen is generally used for short
time, small volume applications (shock freezing) as the frost body builds
up quickly at temperatures of −196°C of the coolant resulting in frozen soil
temperatures between −20°C and −30°C, or even deeper. Brine freezing
uses coolant temperatures of between −30°C and −40°C, allowing frost
body temperatures of −10°C to −20°C to develop. Brine freezing is used in
general for large-volume, long-duration projects since the installation costs
of the refrigerating system are considerable and the time required to achieve
the necessary frost body temperature is longer.
Soil freezing is generally used in water-bearing soils. Ground water low
strongly inluences the development of the frozen soil body and needs to be
considered in the analysis when it reaches values above 1 m/day, as a rule
of thumb. The necessary energy to build up and maintain over time for
the required frozen soil body dimensions can be calculated using the heat
exchange formulae based upon the thermal characteristics of the soil (ther-
mal capacity and conductivity). The design of these dimensions is based
upon structural and heat transfer–related parameters of the frozen and
unfrozen soil. Strength and deformation characteristics of frozen soil are
time dependent and responsible for the creep developing and resulting from
the visco-plastic behaviour of the frozen pore water. The dependency of the
frozen soil strength from temperature and soil characteristics (mainly water
content, density, grain size distribution, and salinity of the ground water)
has been the subject of intensive research in recent years. Recommendations
and special publications (such as from the International Symposium on
Ground Freezing, ISGF) provide details of the method, including the nec-
essary site and laboratory investigations prior to the design and execution
of a soil-freezing project. The thermal and rheological computations to
describe the time and temperature dependent stress strain behaviour of the
frozen soil body are best carried out using the FE method (Kirsch and
Borchert 2008).
Ground water conditions such as seepage low velocity, temperature,
and salinity are important factors of inluence for maintaining the integ-
rity of the frost body dimensions and are often decisive in determining the
distances between the freeze tubes. Their proper knowledge is essential
in optimising the energy cost of any soil freezing project. Quality-control
measures are as with other methods of special foundations; they include
temperature measurements indispensible to guarantee the integrity of the
dimensions of the frost body throughout its maintenance.
It is a well-known phenomenon that freezing of the pore water is accom-
panied by a volume increase of 9%, which results in frost-susceptible soils,
generally of ine contents in excess of 15% in unwanted soil heave. In addi-
tion, the development of ice lenses leads to continuous increase of heave
and/or stresses. This heave is small with suficient surcharge and generally
14 Ground improvement

does not occur in well-draining granular soils where unfrozen water is

expelled by the ice front developing in the pores of the soil.
Soil freezing is a transient ground improvement method leaving behind
after completion of the works very little disturbance in the ground (only
coolant pipes if not extracted) and in this way can be regarded as almost
reversible in restoring original ground conditions, albeit at relatively high
energy cost and with a considerable carbon foot print. After thawing is
completed, ground conditions return to their original state without any
interference with or contamination of soil or aquifer. Access to the method
can be found in special publications such as Jessberger and Jagow-Klaff
(2003), Harris (1995), or Phukan (1985).

REFERENCES

Abelev, Y. M. and Askalonov, V. V. (1957). The stabilisation of foundations for struc-

tures on loess soils. Proc. V. ICSMFE, Paris.
Bell, A. L. (1994). Proc. Conference Grouting in the Ground, Institution of Civil
Engineers, London. Ed. A. L. Bell, Thomas Telford.
Damitio, C. (1970). La consolidation des sols sans cohésion par explosion.
Construction (France), 25:100–8.
EA. (2010). Carbon calculator for construction v 3.1.1. Environment agency UK
spreadsheets and guidance. Jan.
Egan, D. and Slocombe, B. C. (2010). Demonstrating environmental beneits of
ground improvement. Proceedings of the Institution of Civil Engineers,
Ground Improvement, 163(1): 63–9.
Gambin, M. (2004). Densiication des sables lâches par explosifs. In: ASEP-GI 2004,
Vol. 2, 513–540. Ed. Magnan. Presses de l’ENPC/LCPC. Paris.
GEMIS. (2010). Database version 4.6, German Öko Institut for Applied Ecology,
Berlin, Dec.
Glossop, R. (1961). The invention and development of injection processes.
Géotechnique 10:91–100.
Gohl, W. B. , Jefferies, M. G., Howie, J. A. and Diggles, D. (2000). Explosive compac-
tion design, implementation and effectiveness. Géotechnique, 50(6):657–65.
Hall, M. J. (2006). A Guide to Calculating the Carbon Debt and Payback Time for
Wind Farms. London: Renewable Energy Foundation.
Hammond, G. and Jones C. (2011). Inventory of carbon and energy (ICE) version 2.
Database: Sustainable Energy Research Team, University of Bath (UK) January.
Harris, J. S. (1995). Ground Freezing in Practice. Thomas Telford.
Ivanov, P. (1967). Compaction of non-cohesive soils by explosions. US Interior Dept.
Report No. TT-70-57221.
Jessberger, H. L. and Jagow-Klaff, R. (2003). Ground freezing. In: Smotzcyk, U. (ed.)
Geotechnical Engineering Handbook, Vol. 2. 117–168. Ernst & Sohn: Berlin.
Joosten, H. (1926). Verfahren zur Verfestigung von Gebirgsschichten. Deutsches
Reichspatent Nr. 441622.
Kirsch, K. and Kirsch, F. (2010). Ground Improvement by Deep Vibratory Methods,
Spon Press: London and New York.
 Introduction and background 15

Kirsch, F. and Borchert, K.-M. (2008). Bemessung von Vereisungskörpern bei

Sicherungsmassnahmen im innerstädtischen Tunnelbau mit dem Teilsicherheits-
konzept. 23rd Ch. Veder Koll., Schriftenreihe der Gruppe Geotechnik Graz.
Heft 33.
Kolymbas, D. (1992). Sprengungen im Boden. Bautechnik 69(8):424–31.
Littlejohn, G. S. (1993). Chemical grouting. In Ground Improvement, Ed. Moseley.
Blackie: London. 100–29.
Mitchell, J. K. (1981). Soil improvement. State of the art report. Proc. X ICSMFE,
Stockholm.
Phukan, A. (1985). Frozen Ground Engineering. Upper Saddle River, NJ: Prentice
Hall.
Pinske, M. A. (2011). Life cycle assessment of ground improvement methods. MSc
Thesis, University of California, Davis.
Raju, V. R. (1994). Spontane Verlüssigung lockerer granularer Körper-Phänomene,
Ursachen, Vermeidung. Diss. University Fridericiana Karlsruhe.
Smoltczyk, U. (1983). Deep compaction. General Report. Proc. VIII ECSMFE,
Helsinki, Vol. 3.
Solymar, Z. V., et al. (1984). Earth foundation treatment at Jebba Dam site. Journal
of Geotechnical Engineering 110(10):1415–30.
Tamaskovics, N. (2000). Beitrag zu Klärung der Mechanismen von Verdichtun-
gssprengungen. Diss. Technische Universität Bergakademie Freiberg.
Wintzingerode, W., Zöhrer A., Bell, A. L. and Gisselmann, Q. (2011) .Calculations
on greenhouse gas emissions from geotechnical construction processes.
Geotechnik 3:218–21.
Zöhrer, A., Wehr, J. and Stelte, M. (2010). Is ground engineering environmen-
tally friendly? Ecological balance of foundation engineering methods. 11th
International EFFC-DFI conference, session 3: sustainability in the foundation
industry, London, 26 to 28 May.
Chapter 2

Deep vibro techniques

Jimmy Wehr and Wolfgang Sondermann

CONTENTS

2.1 Introduction and history................................................................. 17

2.2 Vibro processes............................................................................... 19
2.3 Vibro plant and equipment ............................................................. 26
2.4 Design and theoretical considerations............................................. 28
2.4.1 Vibro compaction ................................................................ 28
2.4.2 Vibro replacement stone columns ........................................ 31
2.5 Applications and limitations........................................................... 41
2.6 Monitoring and testing................................................................... 42
2.7 Carbon footprint ............................................................................ 44
2.8 Case histories.................................................................................. 46
2.8.1 Vibro compaction for artiicial islands in
Dubai (2001–2008)............................................................. 46
2.8.2 Vibro stone columns for infrastructure works in
Germany (2009) .................................................................. 46
2.8.3 Vibro stone columns for shipyard infrastructure in
India (2009) ......................................................................... 47
2.8.4 Vibro concrete columns for a sewage treatment plant in
Malaysia (2008)................................................................... 49
2.9 Conclusions .................................................................................... 50
References ............................................................................................... 50

2.1 INTRODUCTION AND HISTORY

For over 70 years, depth vibrators have been used to improve the bearing
capacity and settlement characteristics of weak soils. Vibro compaction
is probably the oldest dynamic deep compaction method in existence. It
was introduced and developed to maturity by the Johann Keller Company
in 1936, which enabled the compaction of noncohesive soils to be per-
formed with excellent results. A detailed description of the method from

17
18 Ground improvement

its beginnings up to the pre-war period is given by Schneider (1938) and by

Greenwood (1976) and Kirsch (1993) for the period thereafter.
This original process, now referred to as vibro compaction, has since
been applied successfully on numerous sites around the world. When car-
rying out compaction work using the vibro compaction method in water-
saturated sands with high silt content, these sands, when lowering the depth
vibrator and during subsequent compaction, are liqueied to such an extent
that the compaction effect only occurs after a very long vibration period or
it does not occur at all. In such soils, the vibro compaction method reaches
its technical and economic limits.
In 1956, a technique to insert the vibrator into the soil without the aid
of simultaneously lushing in water was developed to overcome the limi-
tations of the vibro compaction method. After the vibrator is lifted, the
temporarily stable cylindrical cavity is illed with coarse material, section
by section. The coarse material is then compacted by repetitive use of the
vibrator. This vibro replacement procedure came to be known as the con-
ventional dry method. Such technical developments in dense stone column
construction allowed for a greater range of treatable weak natural soils and
man-made ills. Vibro replacement continues to be widely used in Europe
to improve weak soil. It has a reputation for providing stable ground which
allows for safe and economic construction of residential and light commer-
cial and industrial structures.
The conventional dry method utilises the vibrator to displace the sur-
rounding soil laterally rather than for primary compaction of the original
soil. The crushed stone is pressed laterally into the soil during both the
cavity-illing stage and compaction stage. This produces stone columns that
are tightly interlocked with the surrounding soil. Groups of columns cre-
ated in this manner can be used to support large loads. The conventional
dry method reliably produces stone columns to depths of 8 m in cohesive
soils that have a shear strength of at least 20 kN/m 2 .
Bottom feed vibrators, which introduce the stones through the vibra-
tor tip during lift, are used to overcome the disadvantage of possible cav-
ity collapse that can occur with the conventional dry method in cohesive
soils with high water content. During withdrawal of the vibrator, stone and
compressed air are delivered through the vibrator tip, preventing cavity col-
lapse. This method is known as dry vibro replacement. In 1972 this method
was patented in Germany.
Reliable stone column production by vibro compaction in cohesive soils
with high water content is achievable with the aid of a heavy water jet.
Water is jetted from the vibrator tip as the vibrator is lowered to the desired
depth. Mud lushes loosened soil and rises to the surface, stabilising the
cavity. This is known as the wet vibro replacement method.
After the bottom feed system was developed in 1976, it was possible to
install injected stone columns by means of an injection of a cement-bentonite
 Deep vibro techniques 19

suspension near the bottom of the vibrator (Jebe and Bartels 1983). The voids
of the stone column skeleton are thereby illed with this suspension. Finally,
vibro concrete columns were developed using a conventional concrete pump
to deliver the concrete to the bottom of the vibrator via the tremie system.
In very soft nearly liquid soils, vibro replacement is not applicable due
to the lack of lateral support of the soil. A geotextile coating may be used
around the column to ensure ilter stability and to activate tensile forces to
avoid lateral spreading of the column. This method was developed in 1992
and irst applied in early 1993 for a dam project in Austria (Keller 1993).
A compilation of various projects with geotextile-coated columns may be
found in Sidak et al. (2004).
These techniques have been chosen for many major structures in the
United States and Europe, endorsing their value in promoting safe and eco-
nomic foundations to a wide range of buildings and soil conditions. Probably
the oldest recommendation on the use of vibro was issued by the German
transport research society in 1979 (FGFS 1979). Later, the US Department
of Transportation published the manual Design and Construction of Stone
Columns (USDT 1983), followed by the British ICE Speciication for Ground
Treatment (ICE 1987) and the BRE publication Specifying Stone Columns
(BRE 2000). The latest effort has been made by the European community
to standardise the execution of vibro works in Ground Treatment by Deep
Vibration (European Standard EN14731, 2005).

2.2 VIBRO PROCESSES

The operational sequence of the vibro compaction method is illustrated in

Figure 2.1. During operation, the cylindrical, horizontally vibrating depth
vibrator is usually suspended from a crane or like equipment. It weighs
15–40 kN, with a diameter of 30–50 cm and a length of 2–5 m. Details
are provided by Kirsch and Kirsch (2010). The vibrator reaches application
depth by means of extension tubes.
The vibrator shell is constructed of steel pipe, forming a cylinder. Eccentric
weight(s) in the lower section are powered by a motor at the top end of a
vertical shaft within the vibrator. Energy for the motor is supplied through
the extension tubes. The rotational movement of the eccentric weights causes
vibrations of the vibrator. The vibratory energy is transferred from the vibra-
tor casing to the surrounding soil. This energy affects the surrounding soil
without being dependent on the vibrator’s depth of operation. A vibration
damping device between the vibrator and extension tubes prevents the vibra-
tory energy from being transmitted to the extension tubes. Supply pipes for
water and air (optional) are also enclosed in the extension tubes. The pipes can
deliver their payload through the vibrator tip as well as through special areas
of the extension tubes to aid the ground penetration action of the vibrator.
20 Ground improvement

Figure 2.1 Vibro compaction method operating phases. (Courtesy of Keller Group.)

During vibro compaction, the motor runs as the depth vibrator is

inserted into the soil (Figure 2.1). The insertion is aided by water lushing.
Field experience has shown that penetration is more effective when a larger
volume of water is used rather than a higher pressure. The water low will
expel some loosened sand through the annulus around the vibrator. The
granular soil targeted for compaction sees a fast reduction in temporary
excess pore water pressure. At compaction depths greater than 25 m, addi-
tional lushing lines and compressed air may need to be utilised.
The water and air lows are normally stopped or reduced after the vibra-
tor arrives at its speciied depth and the compaction process stages have
been initiated. Field experience has determined that lifting the vibrator in
stages of 0.5 m or 1.0 m after 30–60 seconds of application tends to pro-
duce the best results. During the compaction process, granular material
adjacent to the vibrator sees a reduction in pore volume, which is compen-
sated for by introducing sand via the annulus. It is possible for settlement
of the surface to range from 5%–15% of the compaction depth. This range
depends on the density prior compaction, as well as the targeted degree
of compaction. After the initial insertion and compaction processes have
been completed at a particular location, the vibrator is moved to the next
location and lowered to the depth speciied for compaction.
Compacted soil elements with speciied diameters can be created by
performing the compaction procedures in grid patterns. Open pit brown
coal mining areas, such as those in the eastern part of Germany, have had
vibro compaction performed at depths of 65 m. Typically, the layout of
 Deep vibro techniques 21

compaction probe centres is based on an equilateral triangle. A distance

between 2.5 and 5.0 m usually separates the centres. This distance is
determined by grain crushability (shell content), required density, vibrator
capacity, and grain-size distribution of the sand. The production stage of
extensive projects can be greatly enhanced if a comprehensive soil study is
done, with the added beneit of a test programme prior to going out to ten-
der/bidding on the project. Guideline values for the strength properties of
sand, which can aid the design of such projects, are displayed in Table 2.1.
Currently, depth vibrators are used to produce vibro stone columns in
cohesive soils that exhibit low water content. For this production variant
to be successful, the soil consistency must be able to hold the form of the
entire cavity after the vibrator has been removed. This allows for the sub-
sequent repeated delivery and compaction of stone column material to pro-
ceed uninhibited by obstruction. With the dry or displacement method,
the soil cavity is prevented from collapsing by the compressed air being
released from the vibrator tip.
An alternative method to construct vibro stone columns in cohesive soils
with high water content involves the use of a strong water jet that ejects
water under high pressure from the vibrator tip. The cavity is stabilised by
the mud that rises to the surface and lushes out loosened soil. The cavity
is then illed in stages, through the annulus, with coarse ill, which sur-
rounds the vibrator tip and is compacted into the stone column form as the
vibrator is lifted. This is known as the wet/replacement method. A mud,
or ‘spoil,’ containing high quantities of soil particles is transported to spe-
cially designed settling tanks, or ponds, by way of trenches. This procedure
is complicated and can be messy, but it is important to separate the water
and mud from the operations area, where it is easily accessed when the time
comes to discharge it (Kirsch and Chambosse 1981).

Table 2.1 Guideline values for the strength properties of sand

Very Medium Very
Density loose Loose dense Dense dense
Relative density ID [%] <15 15–35 35–65 65–85 85–100
SPT [N/30 cm] <4 4–10 10–30 30–50 >50
CPT qc [MN/m2] <5 5–10 10–15 15–20 >20
DPT (light) [N/10 cm] <10 10–20 20–30 30–40 >40
DPT (heavy) [N/10 cm] <5 5–10 10–15 15–20 >20
Dry density γd [kN/m3] <14 14–16 16–18 18–20 >20
Modulus of deformation [MN/m2] 15–30 30–50 50–80 80-100 >100
Angle of internal friction [o] <30 30–32.5 32.5–35 35–37.5 >37.5
Source: Kirsch, K. (1979). Geotechnik 1:21–32.
Note: After completion of the vibro compaction work, it may be necessary to re-compact the work-
ing surface down to a depth of about 0.5 m using surface compactors.
22 Ground improvement

Grain diameters of the stones and gravel which comprise the ill material
for the wet method range from 30 to 80 mm. Stone column installation to
depths as great as 43 m has been reported (Wehr 2008). The wet method
guarantees stone column continuity for a wide range of soft soils.
Grain diameters of the stones or gravel that comprise the ill material
when using a bottom feed vibrator typically range from 10–40 mm. The ill
is delivered to the vibrator tip by means of a pipe. After the vibrator arrives
at the speciied depth, compressed air is used to help deliver the ill as the
vibrator is subsequently lifted in stages as it compacts the ill (Figure 2.2).
Carrier equipment typically consists of specially designed machines,
known as vibrocats, which have vertical leaders. The vibrocats control the
complex bottom feed vibrators, equipped with material lock and storage
units, which deliver ill material to the vibrator by means of specialised
mechanical or pneumatic feeding devices (Figure 2.3).
Vibro cats possess a particular feature which is an additional downward
force (Figure 2.4). This so-called ‘activation force’ of approximately 150 kN
causes a better vertical compaction of the column material and a repeated
vertical loading. Furthermore, it is possible to increase the column diameter
easily because the vertical action of the vibrator tip leads to a horizontal
displacement with eficient compaction.
The installation of vibro mortar columns is similar to the dry vibro
replacement method apart from the cement suspension illing the voids of
the stone skeleton inside the column. This results in a much stiffer column
compared to a conventional vibro stone column.
For the installation of vibro concrete columns, the tremie system is con-
nected to a mobile concrete pump. Before penetrating, the system is charged
with concrete. The vibrator then penetrates the soil until the required depth
has been achieved. The founding layer, if granular, is further compacted by
the vibrator. Concrete is pumped out from the base of the tremie at positive
pressure. After raising the vibrator in steps, it re-enters the concrete shaft,
displacing it into a bulb until a set resistance has been achieved. Once the
bulb end is formed, the vibrator is withdrawn at a controlled rate from
the soil while concrete continues to be pumped out at positive pressure.
Once completed, the column can be trimmed and reinforcement placed as
required.
Vibro geotextile columns consist of a sand or stone core with a geotextile
coating. The advantage of a vibro geotextile column to other geotextile
columns (Schüßler 2002) is the well-densiied granular inill resulting only
in small settlements of the soil-column system (Trunk et al. 2004).
The installation is usually performed in several steps in order not
to damage the geotextile. First a hole is created with the vibrator to the
required depth and the vibrator is extracted. In the next step the geotextile
is mounted over the vibrator above the ground surface, and subsequently
the penetration is repeated with the geotextile to the same depth as before.
 Deep vibro techniques 23

Material gate

Vibrator and
pipe extension
(storage container)

Elastic coupling

Pipe for
material
Electric motor

Eccentric
weight

Material outlet

Figure 2.2 Details of a bottom feed vibrator. (Courtesy of Keller Group.)

24 Ground improvement

Figure 2.3 Vibro cat with bottom feed vibrator. (Courtesy of Keller Group.)
 Deep vibro techniques 25

Activation force

Displacement
and
compaction

Figure 2.4 Vibro cat with activation force. (Courtesy of Keller Group.)
26 Ground improvement

On the way up, it is preferable that stones are illed and densiied inside the
geotextile like in the usual dry bottom feed process.
If there is only one certain very soft layer it is possible to irst build a
vibro stone column below this layer, insert a vibro geotextile column or
a vibro mortar column only in the very soft layer for economical reasons,
and inish the upper part of the column as an ordinary vibro stone column.

2.3 VIBRO PLANT AND EQUIPMENT

The equipment developed for the vibro compaction and vibro replacement
processes comprises four basic elements:

1. The vibrator, which is elastically suspended from extension tubes

with air or water jetting systems
2. The crane or base machine, which supports the vibrator and exten-
sion tubes
3. The stone delivery system used in vibro replacement
4. The control and veriication devices

The principal piece of equipment used to achieve compaction is the vibra-

tor (Figure 2.5). The drive mechanism can be an electric motor or a hydrau-
lic motor, with the associated generator or power pack usually positioned
on the crawler rig in the form of a counter weight.
The typical power range in vibrators is 50–150 kW, and can go as high
as 300 kW for the heaviest equipment. Rotational speeds of the eccentric
weights in the cases of electric drives are determined by the frequency of
the current and the polarity of the motor. For example, 3,000 rpm or 1,500
rpm vibrating frequency are obtainable from a 50 Hz power source, and
3,600 rpm or 1,800 rpm vibrating frequency from a 60 Hz power source
with a single or double pole drive, respectively. A 5% reduction in the fre-
quency applied to the ground occurs, corresponding to the magnitude of
the ‘slip’ experienced with asynchronous motors. The use of frequency con-
verters has recently become economical as a result of modern control tech-
nology. The frequency converters enable limited variation of the operating
frequency of the electric motors.
During rotation, the eccentric weight generates horizontal force (see
Figure 2.5). This horizontal force is transmitted to the ground through the
vibrator casing and (depending on the vibrator type) ranges from 150–700
kN. Details are provided by Kirsch and Kirsch (2010). When the vibrator
is freely suspended with a lack of lateral coninement, the vibration width
(double amplitude) totals 10–50 mm. Acceleration values of up to 50 g are
obtainable at the vibrator tip. It is practically complicated to measure cru-
cial operational data during the compaction process. Therefore, any data
 Deep vibro techniques 27

Extension

Elastic coupling

Water
or
air supply

Electric motor

Eccentric
weight

Tip

Figure 2.5 Depth vibrator and principle of vibro compaction. (Courtesy of Keller
Group.)

given on vibrators apply to those which are freely suspended, lacking lateral
coninement.
It is up to the designer to create a vibrator optimal for the speciic appli-
cation. One major challenge of design lies with keeping maintenance costs
within standards that are economically tolerable. Based on ield experience,
the most effective compaction of sands and gravels is done by vibrating fre-
quencies which approach the natural soil-vibrator system frequency, or ‘res-
onance’ for elastic systems, which ranges between 20–30 Hz (Wehr 2005).
Fellin (2000), who considered vibro compaction a ‘plasto-dynamic prob-
lem,’ has conirmed theoretically knowledge gained from practical vibro
operation conditions. Fellin’s goal, by constant analysis of information
obtained on the vibrator movement during compaction performance, was
to create ‘on-line compaction control.’ His work’s theoretical results con-
irm the observation that when using a constant impact force, the vibra-
tion’s effect range increases as the vibrator frequency decreases, whereas
compaction increases when the impact force increases.
The thickness of soil depths to be treated determines the overall length
of vibrator, extension tubes, and lifting equipment, which in turn deter-
mines the size of crane to be used. Purpose-built tracked base machines
28 Ground improvement

(vibrocats) have been constructed to support vibrators: irst, to ensure the

columns are truly vertical, and second, to be able to apply the frequently
required or desired vertical compressive force, which accelerates the intro-
ducing and compacting processes.
The construction of stone columns requires the importation and han-
dling of substantial quantities of granular material. This stone is routinely
handled with front end loaders, working from a stone pile and delivering
stone to each compaction point.
To increase the performance of the vibro system, multiple vibrators may
be applied on one base machine. For example, a barge with a 120–150 t
crane was used for the Seabird project in India with four vibrators (Keller
2002). Alternatively, a special frame was constructed on a barge suspend-
ing ive vibrators (Keller 1997).

2.4 DESIGN AND THEORETICAL CONSIDERATIONS

2.4.1 Vibro compaction

The purpose of vibro compaction is the densiication of the existing soil.
The feasibility of the technique depends mainly on the grain-size distribu-
tion of the soil. The range of soil types treatable by vibro compaction and
vibro replacement are given in Figure 2.6. The degree of improvement will
depend on many more factors including soil conditions, type of equipment,
procedures adopted, and skills of the site staff. Such variables do not permit

Transition
Clay Silt Sand zone Gravel
100 100

80 80

60 Vibro-replacement 60
Passing by weight [%]

Vibro-compaction
40 40

20 20

0 0
0,002 0,006 0,02 0,06 0,2 0,6 2,0 6,0 20 60
Particle size [mm]

Figure 2.6 Range of soil types treatable by vibro compaction and vibro replacement
(stone columns).
 Deep vibro techniques 29

a/sin 60º
a

(1 + b  a

(1 + 2b  a
3 × a/3

b = 0.2 – 0.4
Compaction probe
Pre - penetration test
Post - penetration test

Figure 2.7 Trial arrangement for vibro compaction. (From Moseley, M.P. and Priebe, H.J.
(1993). In: Moseley, M.P. (ed.) Ground Improvement, London, England: Blackie
Academic & Professional.)

an optimum design to be established in advance but rather require the exer-

cise of experience and judgement for their successful resolution.
For small projects, the design of vibro compaction work can be based
on the experience of the contractor. For large projects it is preferable and
advisable to conduct a trial in advance of contract works. A typical lay-
out of vibro compaction probes for a trial is given in Figure 2.7. The trial
allows for three sets of spacings between probes, together with pre- and
post-compaction testing, often performed using cone penetration testing
equipment. The degree of improvement achieved can be used to optimise
the design, as shown in Figure 2.8.
The technical success of vibro compaction work is measured by the level
of densiication achieved against a speciied target. The densiication can be
readily checked using standard penetration tests or, preferably, cone pene-
tration tests. Comparisons can be made between pre- and post-compaction
testing, and care should be taken to ensure that the same techniques of test-
ing are used in each situation. Control of performance is a further impor-
tant element in carrying out vibro compaction work. This is best achieved
30 Ground improvement

MN/m2 Cone resistance qs

25

20

15 Sand

10

Initial value

5
0.06 2.0
Grid area A
0
0 1 2 3 4 5 6 7 8 9 10 m2

Figure 2.8 Results of vibro compaction trial. (From Moseley, M.P. and Priebe, H.J. (1993).
In: Moseley, M.P. (ed.) Ground Improvement, London, England: Blackie Academic
& Professional.)

by using a standardised procedure, established at the precontract trial, such

as predetermined lifts of the vibrator at predetermined time intervals and/
or predetermined power consumptions. Only such a regular procedure can
reveal whether variations in test results are due to the inherent inhomo-
geneity of the soils being treated or by insuficient compaction.
The soil being treated, the degree of densiication required, the type of
vibrator being used, and production rates all have an inluence on the spac-
ing of vibro compaction probes. Areas treated per probe vary commonly
between 6–20 square metres. Vibrator development over the past decade
has allowed considerable increases in the area treated by each insertion of
the vibrator. This development continues and will enable further expansion
of the treatment envelope.
Sands and gravels bearing negligible cohesion are compatible with vibro
compaction. The silt (grain size <0.06 mm) percentage of such soils should
be less than 5% for ideal performance. Compaction is substantially hin-
dered by clay particles (grain size <0.002 mm) to the point that the pro-
cedure is unable to be performed without extra measures, including the
introduction of coarse-grained ill. Reference to the grain-size distribu-
tion diagram (Figure 2.6) usually determines application limits. However,
application limits for material that is very coarse are typically determined
empirically, taking into consideration the penetration effectiveness of the
respective vibrator. Static cone penetration tests can also serve to estimate
 Deep vibro techniques 31

values of soil compatibility for compaction methods. Given that the local
skin friction-to-point resistance (friction ratio) falls between 0 and 1 and
the point resistance is a minimum of 3 MPa, the soil can be considered to
be compatible (Massarsch 1994).
The eficiency of compaction is also greatly inluenced by the perme-
ability of the soil. When permeability is too low (<10 –5 m/s), compaction
effectiveness decreases as permeability decreases, whereas when permeabil-
ity is too high (>10 –2 m/s), penetration of the soil by the vibrator becomes
increasingly more dificult as the permeability increases (Greenwood and
Kirsch 1983).
The carbonate or shell content is important for the densiication of
highly compressible soils with low cone resistance and high friction ratio.
Cemented soils are not considered here.
Correlations between the CPT cone resistance and the relative density
are well established for silica sand. Unfortunately there are not many refer-
ences concerning this correlation for calcareous sands. Vesic (1965) added
10% of shells to quartz sand, which resulted in a decrease of the CPT cone
resistance by a factor of 2.3. Bellotti and Jamiolkowski (1991) compared
CPT cone resistances qc (silica)/qc (shells) = 1 + 0.015(Dr-20) yielding ratios
between 1.3 and 2.2 increasing with relative density Dr. Almeida et al.
(1992) compared normalised CPT cone resistances of calcareous Quiou
sand and silica Ticino sand, which yielded ratios from 1.8–2.2 proportion-
ate to increasing relative density. Foray et al. (1999) compared pressure-
meter limit pressure of silica sand and carbonate sands, which resulted in
ratios ranging from 2–3 proportionate to increasing initial vertical stress.
Finally Cudmani (2001) looked at normalised cone resistances of seven
sands yielding ratios between 1.4–3.5 depending on initial soil pressure
and relative density. Meier (2009) executed systematic calibration chamber
tests to investigate the inluence of different silica/carbonate sand ratios
including the inluence of coarse material. This concept has been applied to
the Palm Island projects in Dubai by Wehr (2005a).

2.4.2 Vibro replacement stone columns

The reduction of consolidation time and compressibility, and the increase
of load-bearing capacity and shear strength, determine the effect of vibro
replacement in soft ine-grained soils. The in-situ soil characteristics, the
placement and geometry of the stone columns, and the soil-mechanical
properties of the column composition are what determine the scale of
ground improvement achieved. Aside from the settlement rate increase
(generated by the stone column’s drainage effect), the reduction of over-
all settlement is the goal of the vibro column installation. Quite simply,
stone columns are effective in reducing settlement since they are stiffer than
the surrounding soil. Between stone columns and the ground, the effective
32 Ground improvement

stiffness ratio relies considerably on lateral support provided by the sur-

rounding soil when the stone columns have loads put upon them. In order
to mobilise the lateral support and generate the interaction between the soil
and columns, a horizontal deformation is required. This deformation inevi-
tably causes settlement at the ground surface. Bell (1915) relays the most
simplistic relationship for calculating load-bearing behaviour. A maximum
lateral support of σh = γz + 2 cu can be provided by the adjacent cohesive soil
possessing a cohesion cu at depth z. If it is assumed that the passive earth
pressure coeficient Kp = tan 2 (π/4 + ϕ/2) is used, then the above supporting
pressure allows a maximum vertical column stress of σo = Kp (γz + cu), with
ϕ being the angle of the internal friction of the column material (Figure 2.9).
This equation, while underestimating the column’s load-bearing capacity,
still conveys the signiicance of column and ground interaction. The equa-
tion also reveals the differences in load-bearing behaviours of stone col-
umns when compared to load-carrying elements of greater stiffness.
Minimum shear strength of ground proposed for improvement used to
be frequently given in the form of a very conservative cu value of 15 kN/m 2
(AUFS 1979, Smoltczyk and Hilmer 1994). It must be noted that no atten-
tion is given to the positive effects of the three-dimensional behaviour, the
inluences of adjacent columns, the dilatation of column material (Van
Impe and Madhav 1992), and most importantly, the rapid increase in the
soil’s shear strength owed to the stone column’s drainage effect. As a conse-
quence of these matters, the successful production of foundations in much
softer soils with cu ≥ 4 kN/m 2 via vibro replacement has been achievable
(Raju and Hoffmann 1996, Wehr 2006). Many model tests have been con-
ducted in order to more clearly grasp the column/soil interactions and the
inluences of adjacent columns (Hu 1995). In qualitative terms, these tests
show the failure mechanism on one side and the group effect on the other
(Figure 2.10).
With ultimate vertical load, the failure of stone columns is a result of
relatively low lateral support in the upper third (bulging), or the column toe

Shear strength of
Shear strength stone material
τ of soil

σ0

Z
σh
Cu
σ
γZ σh Kp  σh = σ0
ϕ

Figure 2.9 Inluence of lateral support on column stress. (From Brauns, J. (1978).
Bautechnik 55(8):263–271.)
 Deep vibro techniques 33

Figure 2.10 Failure mechanism of vibro replacement stone columns in the case of group
effect. (From Hu, W. (1995). Physical Modelling of Group Behaviour of Stone
Column Foundations, PhD Thesis, University of Glasgow.)

being punched into the underlying soil, such as with ‘loating’ foundations
(Figure 2.11). However, such high rates of deformation precede the failure
in every case that the column’s serviceability is generally no longer pro-
vided. Therefore, we can conclude that the equations used to calculate the
deformation, or ‘serviceability state’ of the discussed foundation, are much
more relevant than the outcome of limit load assessment of stone columns.
Soyez (1987) and Bergado et al. (1994) have conducted a thorough over-
view of the various design methods. The authors show the distinction
between calculating single columns and calculating column grid patterns.
In Europe, Priebe’s (1995) design method for vibro replacement stone col-
umns has gained acceptance as a valid method (Figure 2.12).
Thus, in Figure 2.12, the improvement factor, depending on angles of
internal friction of the stone column, is related to the ratio of the stone
column area and the area being treated by the column. The improvement
factor indicates how many times the compression modulus increases for a
grid of stone columns and to what extent the settlement of a raft founda-
tion will be reduced. Angles of internal friction are usually higher than 45°
(Herle et al. 2008).
34 Ground improvement

P P

Figure 2.11 Failure mechanism of vibro replacement stone columns under vertical load.
(From Brauns, J. (1978). Bautechnik 55(8):263–271.)

The basic design curves assume the stone column material to be incom-
pressible, and Figure 2.13 allows an adjustment to be made for this by
plotting a ictitious area ratio, which has to be added to the actual area
ratio, against the compression modulus ratio for soil and stone column
material.

5
ϕc = 45.0°
Improvement factor n

ϕc = 42.5° µs = 1/3
4
ϕc = 40.0°
ϕc = 37.5°
3
ϕc = 35.0°

1
1 2 3 4 5 6 7 8 9 10
Area ratio A/Ac

Figure 2.12 Design diagram for improving the ground by vibro replacement stone col-
umns. (From Priebe, H.J. (1995). Ground Engineering, December, pp. 31–37.)
 Deep vibro techniques 35

2.0

ϕc = 45.0°
Addition to the area ratio ∆ (A/Ac)

1.6
ϕc = 42.5° µs = 1/3

ϕc = 40.0°
1.2
ϕc = 37.5°

ϕc = 35.0°
0.8

0.4

0.0
1 2 3 4 6 8 10 20 30 40 60 80 100
Constrained modulus ratio Dc /Ds

Figure 2.13 Area ratio addition. (From Priebe, H.J. (1995). Ground Engineering, December,
pp. 31–37.)

With regard to settlement performance, theoretical approaches predomi-

nately refer to an ininite grid of columns. Load tests executed in practice
on footings resting on small numbers of columns do not fulill the assump-
tions. Accordingly, evaluations of settlement performance of a footing on a
limited number of stone columns are only approximations.
Practical design charts that consider load distribution as well as reduced
lateral support on columns situated underneath footing edges have been
presented by Priebe (1995). These charts allow the estimation of settlement
of a rigid foundation on a limited number of stone columns as a function of
the settlement of an ininite raft supported by an ininite grid of columns,
as outlined above.
The method presupposes that the footing area attributed to a stone col-
umn and the foundation pressure are identical. There exists an optimum
layout for a given number of stone columns beneath a footing. However, in
practical applications it is suficient to determine the grid size required for
the calculation by dividing the footing area by the number of columns. The
main chart to use in the evaluation of load tests is shown in Figure 2.14.
The application is relatively simple as the relevant settlement ratio depends
on the number and diameter of the stone columns together with the treat-
ment depth considered.
The United States has seen wider use of Goughnour and Bayuk’s (1979)
iteration method, even though it is generally considered much more com-
plex. A great number of these calculations are derived from empirical or
36 Ground improvement

0.8
Settlement ratio s/s∞

1600
900
0.6

No. of stone columns

400
225
0.4 100
64
36
0.2 16
9
4
1
0
0 4 8 12 16 20 24 28 32
Depth/Diameter ratio d/D

Figure 2.14 Settlement evaluation for isolated footings. (From Priebe, H.J. (1995). Ground
Engineering, December, pp. 31–37.)

semi-empirical equations involving simplifying assumptions that do not

effectively address the deformation behaviour’s intricacies. There is cur-
rently a lack of an acceptable design method to adequately account for all
mechanisms that take part in the load transfer process and is also simple
enough for practical use. Therefore, it is best, before making inal decisions
for the execution of designs for sizeable ground improvement projects, to
install test columns and use the achievable column diameters with the load
test results to ensure an effective outcome (Chambosse and Kirsch 1995).
When determining the stress/deformation behaviour in the service load
range, simulation calculations, such as the inite element method (FEM)
often used in construction, are known to be highly effective. As for vibro
replacement stone columns for ground improvement, Schweiger has pro-
posed a method that utilises a homogenised model, the so-called ‘ground/
column matrix’ (Schweiger 1990). Wehr (2004, 2006a) has produced note-
worthy results regarding the simulation of the failure mechanisms of stone
columns by use of his calculations for single columns and column groups
for rigid and lexible footings. Brauns’ proposed failure modes (1980) and
Hu’s (1995) model tests (Figures 2.11 and 2.10, respectively) have since
been recalculated, resulting in the conirmation that shear zones dictate
the settlement behaviour of columns. The inluence of the different design
parameters has been intensively investigated by Kirsch (2004). Currently,
numerical analysis by means of FEM has gained acceptance as a valuable
tool in designing stone column ground improvement when large projects
are in the design phase or current concepts need optimisation.
 Deep vibro techniques 37

The phenomenon of liquefaction of granular deposits during an earth-

quake has been well documented and increasingly studied. Engineering
opinion has agreed that the role of liquefaction can be minimised by den-
sifying the soils beyond their liquefaction potential for the site-speciic
design earthquake. A second method of minimising the role of liquefac-
tion requires the provision of drainage paths, thus allowing rapid dis-
sipation of pore pressure induced by an earthquake. The inluence of the
drainage capabilities of stone columns have been studied by Baez (1995),
concluding that the Seed and Booker (1976) model is useful if allowable
maximum pore pressure ratios are maintained below 0.6. Further inves-
tigations of in-situ stone columns composition indicated that in sands the
columns generally have an 80/20 proportion (gravel to sand) due to the
installation process.
A combination of vibro replacement and vibro compaction, where dense
permeable stone columns are constructed and the density of the surround-
ing granular soil is increased, provides an excellent solution to liquefaction
problems. Since its irst application at Santa Barbara, California, in 1974
(Engelhardt and Golding 1975), it has been used many times. Perhaps the
most signiicant are the documentations of the performance of the Santa
Barbara project (Mitchell and Huber 1983) following a seismic event which
induced ground accelerations equal to the design earthquake and the study
of 15 sites in the San Francisco area (Mitchell and Wentz 1991) following
the Loma Prieta earthquake. In the latter study, the sites treated by vibro
techniques and the buildings founded on them were shown to have suffered
no damage during the Loma Prieta earthquake.
The acceleration rates that affect the soil in the immediate vicinity of the
depth vibrator greatly surpass those experienced in seismic events. On one
project, peak ground accelerations of 1.7 g were detected 0.9 m from the
stone column’s centre (Baez and Martin 1992). As acceleration increases,
the soil’s shear strength is reduced. In saturated sand, complete liquefaction
is possible in the event that the increase in pore water pressure generated
by the vibrations surpasses the decrease in pore pressure which is naturally
caused by iltration/dissipation (Greenwood and Kirsch 1983).
As long as the treatment medium consists of uniform coarse-grained
sands and gravels with a minimum relative density of 80%, the following
are attainable: acceptable load-bearing capacities, marginal settlement risk,
and assurance against liquefaction induced by seismic events (Smoltczyk
and Hilmer 1994). As the percentage of ines increases, higher densities
become increasingly dificult to achieve. Therefore, when working in uni-
form ine-grained or silty sands, it is beneicial to install stone columns that
enhance the drainage capacity. Cohesive soils appear to be more resistant
to liquefaction than clean sands, but liquefaction is possible as well under
seismic action of relatively long duration and high intensity. Many detailed
site examples are given by Perlea (2000).
38 Ground improvement

It is not possible to estimate, by statistical analyses, the extent to which

the risk of liquefaction is reduced by vibro replacement. The key question
is which part of the forces exerted by an earthquake is borne by the col-
umns without any damages. The simple procedure for the design of vibro
replacement by Priebe (1995) was modiied to account for short-term seis-
mic events (Priebe 1998). In this case it is more realistic to consider defor-
mations of the soil with the volume remaining constant; that is, to calculate
with a Poisson’s ratio of 0.5, which also simpliies the formulae. In the
above-mentioned procedure the improvement factor n 0, which is the basic
value of improvement by vibro replacement, is determined initially using

Ac  1 
n0 = 1 + ⋅ − 1
A  K ac ⋅ (1 − A c A) 

and

Kac = tan 2(45º – φc /2)

where
A = attributable area within the compaction grid
A c = cross-section of stone columns
ϕc = friction angle of column material.

The reciprocal value of this improvement factor is merely the ratio

between the remaining stress on the soil between the columns ps, and the
total overburden pressure p taken as being uniformly distributed without
soil improvement and, as such, can be used as a reduction factor α = 1/n 0.
On the understanding that the loads taken by the columns from both the
structure and the soil do not contribute to liquefaction, it is proposed to
use this factor to reduce the seismic stress ratio created by an earthquake
and hence evaluate the remaining liquefaction potential according to Seed
et al. (1983). A similar approach was proposed by Baez (1995) substituting
the above Kac with a ratio between the shear modulus of the soil and the
stone column.
It is important to mention that excess pore water pressures play an impor-
tant role in reducing the effective stresses but are neglected in the conven-
tional above-mentioned design. A novel liquefaction approach including
pore water pressures was applied by Cudmani et al. (2003) to two sites, one
of them being Treasure Island inluenced by the 1989 Loma Prieta earth-
quake. Liquefaction was predicted in a concentrated zone comprising both
the bottom of a ine sand top layer and an underlying upper part of a silty
sand layer. Mitchell and Wentz (1991) reported on the medical building in
Treasure Island where the upper ine sand layer was improved with stone
columns to a depth of 6.5 m, leaving the lower layer unimproved. This
 Deep vibro techniques 39

resulted in no liquefaction of the improved soil block but in liquefaction

of the silty sand layer below 6.5 m, which was proved by the observation
that the bottom 2.5 m of the 6.5-m-deep elevator shafts drilled prior to the
earthquake were illed with silty sand. Furthermore, sand boils were clearly
visible outside the improved area.
Another aspect is the design of the extent of soil improvement against
liquefaction. An overview is given in Japanese Geotechnical Society (JGS
1998). There are two basic questions to be answered about the necessary
width and depth of the soil improvement outside the loaded area.
Pore water pressures are transmitted from the liqueied area into the
improved area of the ground. It is recommended (JGS 1998) to improve
a lateral area corresponding to an angle of 30 degrees against the vertical
axis starting from the edge of the foundation (point A in Figure 2.15). This
shall be executed down to a nonliqueiable layer. The area ACD in the same
Figure 2.15 exhibited particular unstable behaviour during model tests, and
hence this part should be treated as liqueied in the soil improvement design.
Design guidelines for oil tanks in Japan (JGS 1998) recommend improv-
ing an area adjacent to the footing corresponding to 2/3 of the soil improve-
ment depth, Figure 2.16. Recent research on sand drains for liquefaction
remediation yields the lateral extent to be taken as the liqueiable depth
(Brennan and Madabhushi 2002).
In many design codes and standards, a maximum treatment depth
between 15–20 m is given according to experience. A special design chart

Compacted part of Uncompacted

ground ground

A D

Liquefaction
30°

B C
Impermeable non-liquefiable layer

Figure 2.15 Stabilised area (ABCD) adjacent to foundation. (From Japanese Geotechnical
Society. (1998). Remedial Measures against Soil Liquefaction, from Investigation
and Design to Implementation. Rotterdam, Netherlands: A. A. Balkema.)
40 Ground improvement

2
Oil tank L= 3
1

5 m ≤ L ≤ 10 m

Liquefaction Soil improvement area Liquefaction

I
L

Figure 2.16 Soil improvement area for oil tanks. (From Japanese Geotechnical Society.
(1998). Remedial Measures against Soil Liquefaction, from Investigation and
Design to Implementation. Rotterdam, Netherlands: A. A. Balkema.)

is available for light-weight and small-scale structures to improve a limited

depth leaving a liqueiable soil layer below (JGS 1998).
The time-dependent behaviour of sand or gravel drains may be analysed
using charts proposed by Balaam and Booker (1981). This is an extension
of the Barron solution for excess pore water pressure using the approximate
diffusion theory for consolidation based on Biot’s equation of consolida-
tion. The rates of consolidation are presented in charts depending on the
diameter ratio of the unit cell and the column de /d, the stiffness ratios of the
column and the soil under drained conditions E1/E 2 and a Poisson’s ratio of
0.3, which is assumed equal for the soil and column.
The design method of vibro replacement by Priebe (1995) was extended
by Raithel and Kempfert (2000) to account for tensile hoop forces in a
geotextile which may be used around the columns in soils with cu < 4 kPa.
The hoop force is transformed into a horizontal stress which supports the
column additionally to the soil.
Vibro mortar columns (VMC) and vibro concrete columns (VCC) are
ideal for weak alluvial soils such as peats and soft clays overlying compe-
tent founding strata such as sand, gravels, and soft rock. Working loads
of up to 1,000 kN can be achieved in appropriate soils. The ‘bulb end’
and frictional components of the VCC enable high safe working loads to
be developed at shallower depths than alternative piling systems and thus
generally provide a more economical solution.
The Priebe model to design vibro replacement was extended to allow
also for stiff columns: if the load is higher than the inner strength of the
columns, the conventional vibro replacement design by Priebe (1995) is
executed. But if the column load is lower than the inner strength of the
 Deep vibro techniques 41

columns, the calculation is modiied (Priebe 2003). At irst the settlement

of the soil below the bottom of the VCC is determined using the stress,
which corresponds to the one of a shallow foundation in a homogeneous
half space. This formulation is not on the safe side as the load distribution
is smaller than in homogenous soil due to stiffer vibro columns. In a second
step the settlement is determined from the difference to the increased stress
below the bottom of the columns. This yields the punching effect of the col-
umn toe into the soil below. Because of the difference of the averaged stress,
which has been assumed to be quite small, a certain compensation is given.
The value determined as column punching has to be added to the settle-
ment of the soil below the columns. A similar model has been developed by
Tomlinson (1980) for piled raft foundations.
Once completed, the columns exhibit stiffness 10–20 times greater than
the adjacent soil. Construction of a supplementary layer of compacted
material over the column heads is often performed in order to focus the sur-
face load on the columns. The surface load is focused by means of an arch-
ing effect that occurs as this layer thickens. An alternative method involves
using a horizontal geotextile. Suspended between the column heads, it pre-
vents the columns from puncturing an attenuated load distribution layer
(Kempfert 1995, Sondermann and Jebe 1996, Topolnicki 1996). For vibro
stone columns, such a load distribution layer is not necessary.

2.5 APPLICATIONS AND LIMITATIONS

Vibro compaction is used to increase the bearing capacity of foundations

and to reduce their settlements. Another application is the densiication of
sand for liquefaction mitigation. By reducing the amount of water which
has to be pumped during groundwater lowering, sand can be compacted,
which reduces the permeability. This solution is also possible for dams.
Vibro compaction is limited by the ines and carbonate content (see Section
2.4). Furthermore, a certain distance should be kept to existing buildings
in order to limit settlements of new buildings (Achmus et al. 2010). Depths
down to 65 m have been improved so far by vibro compaction.
Various ways of creating vibro stone columns have been developed in
order to enhance load-bearing capacities of weak soils and limit settlement.
For the support of individual or strip foundations, small groups of col-
umns are employed. Large column grids are placed beneath rigid foun-
dation slabs or load conigurations that exhibit lexibility, as is the case
with storage tanks and embankments. Due to inherent higher shear resis-
tance, vibro replacement columns are a good choice for the enhancement
of slope stability. When drainage takes precedence over bearing capacity,
vibro drain columns can be employed to function as drains. This drainage
type of sand or stone column is constructed simply by lifting the vibrator
42 Ground improvement

without compaction, leaving the sand or stone in a state ranging from loose
to medium dense (European Standard EN 15237, 2007).
Vibro stone columns are not suitable in liquid soils with a very low und-
rained cohesion because the lateral support is too small. However, vibro
stone columns have been installed successfully in soil with 5 kPa < cu < 15
kPa, see Section 2.7. In case of very hard and/or cemented layers (i.e., cap-
rock) or very well-compacted surface layers, pre-boring may be necessary
to assist the penetration of the vibrator. Concerning the distance to build-
ings the same applies as for vibro compaction (Achmus et al. 2010). Depths
down to 43 m have been improved so far by vibro replacement.

2.6 MONITORING AND TESTING

Part of the state of the art methods is to monitor and record in great detail
the operating parameters of any deep vibro work. Details are given in
the European Standard ‘Ground treatment by deep vibration’ (European
Standard EN 14731, 2005).
Vibro compaction is monitored online with devices that record, as a
function of time, penetration depth, energy consumption of the motor and,
if necessary, pressure and quantity of the lushing media used. If the vibra-
tor frequency can be adjusted during the compaction process, this param-
eter is also recorded.
For the vibro replacement method, all of the essential parameters of
the production process (depth, up/down speed, activation force, vibrator
energy, and stone/concrete consumption) are recorded continuously as a
function of time, providing the user with visible and controllable data for
producing a continuous stone column. A typical printout for stone column
construction is given in Figure 2.17. Monitoring of the activation force is
important to check that there is suficient vertical compaction and hori-
zontal displacement of the column material. Additionally, the monitoring
of the ill level in the material supply tube indicates exactly how much
material is inserted per linear metre column length. Monitoring the acti-
vation force and the ill level together guarantees a high-quality column.
Such instrumentation is available for leader-mounted, bottom feed vibra-
tor systems and has been used in Europe since the 1980s (Slocombe and
Moseley 1991).
Vibro compaction and vibro replacement are increasingly evaluated by
means of 2D and 3D plots whereby areas of inhomogeneous soil can be
detected. To monitor the installation quality covering a large area, the
programme ‘VibroScan’ visualises automatically recorded vibro data, like
depth, compaction energy, and activation force. Borelogs from core drill-
ings, cone penetration tests, and standard penetration tests (SPT) represent
only punctual explanations, which can never show the entire ground. After
 Deep vibro techniques 43

Process:
Inventar: 9130744 Site: 331297
Lot: 96 Point: 156 Ref. No.: 13
Date: 08.04.03 Time: 13:23:08 Interval: 2 sec
3
Weight: 16.0 kN/m Dep.: Consulting and development
Legende:

Time Depth Up/Dowm Force Current Fill level Frequency Air

[sec] Speed [m] [Hz] pressure
[m] [m/min.] [kN] [A] [bar]

0 5 10 –50 0 50150 0 150 0 200 0 5 10 0 30 60 0 0.5 1.0

0

200

400

600

800

1000

1200
Total time: 19.85 min Col. Length: 7.3 m Depth 7.3 m

Figure 2.17 Typical printout for stone column construction. (Courtesy of Keller Group.)

collecting the data of all compaction points with VibroScan, a planar over-
view of the entire area of the ground improvement is displayed. Such areas
are highlighted whose soil conditions deviate from the remaining area, and
therefore a deviating quality of the ground after ground improvement mea-
sures is shown. The execution grid can be adapted to the respective soil
44 Ground improvement

conditions, and auxiliary points can be arranged in areas of worse soil

quality in order to receive a preferably constant quality and thus an opti-
mum homogenisation of the improved ground at the end.
In addition to the online control, the inal site records include the posi-
tion and elevation of columns; the source, type, and quality of imported
material; and, if necessary, environmental factors (noise, vibrations, etc.).
Cone penetration tests and SPTs are commonly used to verify the success
of vibro compaction, with CPT being the better of the two. To compare
the initial and inal compressibility of the soil, pre- and post-tests undergo
comparison. When evaluating post–deep compaction work test results, the
ageing effect must be taken into account. This ageing effect on strength
goes on for up to several weeks after the column has been installed. Many
projects have demonstrated that the strength of compacted sands has the
potential, over several weeks, to increase anywhere from 50%–100%.
This substantial increase is attributed to pore water pressure reduction,
sometimes in combination with the re-establishment of physical and chemi-
cal bonding forces to the column’s grain structure (Mitchell et al. 1984,
Schmertmann 1991, Massarsch 1991). Taking this strength increase into
account, it is best to wait at least one week after compaction work before
conducting formal compaction tests. The technical literature record con-
tains much information regarding reports on compaction tests and moni-
toring (Covil et al. 1997, Slocombe et al. 2000).
The performance of vibro stone columns is monitored only for large proj-
ects using large plate load tests, which should be carried out by loading
a rigid plate or cast in-situ concrete pad big enough to span one or more
columns and the intervening ground. Zone load tests should be carried out
by loading a large area of treated ground, usually by constructing and load-
ing a full size foundation or placing earth ill to simulate widespread loads.
In case of soils with high sensitivity S > 8, the soil structure in the vicinity
of the vibro stone columns may be disturbed resulting in a decrease of the
initial undrained shear strength. It may take up to several months of recov-
ery to reach the initial value. Enough time should therefore be foreseen
between the installation of the columns and the tests.

2.7 CARBON FOOTPRINT

Beginning with the ratiication of the Kyoto Protocol in 1997, the awareness
of the global warming effect has risen all over the world. Approximately
1/5 of global greenhouse gas emissions are associated with the manufac-
turing and construction industries. In recent years, the awareness of the
carbon footprint has reached the construction industry, leading to sev-
eral certiicates for buildings and engineering structures. The foundation
of a building can have a signiicant impact on the total carbon footprint,
 Deep vibro techniques 45

depending on the local ground conditions. Ground improvement technol-

ogies can signiicantly reduce greenhouse gas emissions because the raw
materials used have by far the biggest impact on the total emission of the
foundation works, followed by the diesel consumption of the machinery
used. By improving the existing soil, the use of concrete can be minimised
or even avoided, resulting in a signiicant reduction of the foundation
work’s carbon footprint, and hence that of the total structure. Taking into
account that concrete has a high input on carbon footprint emissions due
to the manufacturing process, ground improvement technologies as vibro
compaction or vibro stone columns are a sustainable and environmentally
friendly solution.
Figure 2.18 shows the computed CO2 equivalents in kg per linear m for
different special foundation methods (Zöhrer et al. 2010). The vibro stone
column method with gravel possesses the environmental friendly small-
est CO2 equivalent per linear m foundation element, followed by ready
mixed mortar columns (VMC) and VCC. The large amount of cement
used in CFA piles, deep soil mixing (DSM), and especially in bored piles
(BP) leads to high or very high CO2 equivalents. For foundation or ground
improvement elements using binders, the employment of Portland cement
instead of blast furnace cement leads to approximately 60%–70% higher
CO2 output.

900
750-800
800
660-690
700
kg CO2 - eq./m

600
500

400
300 243-247
200 134-141139-143
70-73 79-82 92-96
100
7-10 10-12 42-45 55-58
0
VSC VSC VMC VCC VMC CF VCC CFA DSM DSM BP BP
with with with with with ∅ with ∅ with with ∅ ∅
gravel ballast BC BC PC 45 cm PC 64 cm BC PC 60 cm 90 cm

Figure 2.18 CO2 equivalents for ground improvement and/or deep foundation meth-
ods. VSC = vibro stone columns, VMC = vibro mortar columns, VCC =
vibro concrete columns, CFA = continuous light auger piles, DSM = deep
soil mixing, BP = bored piles, PC = Portland cement, BC = blast furnace
cement. (From Zöhrer, A., Wehr, W., and Stelte, M. (2010). Is ground engi-
neering environmentally friendly? Proceedings of the 11th International EFFC-
DFI Conference, Session 3: Sustainability in the Foundation Industry, May 26–28,
2010, London, England.)
46 Ground improvement

2.8 CASE HISTORIES

2.8.1 Vibro compaction for artificial

islands in Dubai (2001–2008)
Extensive ground improvement using vibro compaction of reclaimed
sand ill was carried out between 2001 and 2008 in Dubai for the ‘Dubai
Waterfront’ project ‘’with a total area of 130 km2, consisting out of the
‘Palm Jumeirah Island’, the ‘Palm Jebel Ali’, the ‘Palm Deira’, and the
‘World Island’ (Wehr 2005a; Haß et al. 2010).
The material used to reclaim the islands originated from the seabed in
front of the Dubai coast. The material was then put in place by means of a
dredging process, leading to a loose state of density. Therefore, all that soil
had to be compacted down to the bearing layer by means of deep vibro com-
paction. Design criteria allowed maximum settlements of 25 mm and a resis-
tance against liquefaction in case of an earthquake with a magnitude M =
6.0. The allowed maximum in angular rotation was limited to 1:500. The
design loads were up to 150 kN/m2. Frequently, crane-hung twin vibrators
were used for the vibro compaction works. To proof the achievement of the
required compaction, a CPT reference diagram has been developed by means
of Eurocode 8. Due to the fact that the material used had a high carbonate
content originating from seashells and corals, the results of CPT soundings
had to be adjusted to consider the modiied strength of the reclaimed sand ill
in comparison with silica sand. This shell correction factor f = qc (silica sand)/
qc (shell sand) equalled 1.3. Quality control was executed by means of cone
penetration tests every 900 m2 of improved ground and material extractions
by means of borings every 20,000 m2. An area-wide geodetic levelling was
also carried out to establish the settlements originating from the compaction
processes. The individual compaction points were monitored by automatic
logging devices attached on the crane units. Every 2 km2, a static load test
was carried out. These tests conirmed the compaction success and the theo-
retical assumptions that have been made in the beginning (Figure 2.19).

2.8.2 Vibro stone columns for infrastructure

works in Germany (2009)
Vibro stone columns have been carried out for roads and bridges at the new
Berlin Brandenburg International Airport. Sandy boulder clay was improved
down to a depth of 8.0 m below ground surface, covering a total area of
90,000 m2. The aim was to reduce the settlement behaviour of the soil by fac-
tor 2. The bearing layer has been identiied as being sand with cone penetra-
tion resistance of qc > 10 MN/m2. The spacing and the vibro stone column’s
diameter based on results of ield trials and static load tests carried out in
representative locations with unfavourable soil conditions. These locations
have been identiied by additional CPTs prior to the execution of works. In
 Deep vibro techniques 47

Figure 2.19 Compaction works at Palm Island.

addition to these ield trials and load tests, numerical calculations have been
executed to examine additional soil improvement effects by means of stiffness
improvement of the soil surrounding the columns and stiffness improvement
by means of displacement effects between columns. The static load tests and
the numerical calculations resulted in a required stiffness modulus of 21 MN/
m2 and a stone column pattern of 2.75 m × 2.75 m. To be on a safe side, an
execution point distance pattern of 2.5 m × 2.5 m was chosen. For some
single footings, a reduction of the point distance to a smaller pattern was nec-
essary to meet the settlement criteria. All compaction points were identiied
by means of GPS technology. The friction angle of the stone columns, used in
the numerical calculations was obtained by means of laboratory shear tests
prior to the design works, resulted in a friction angle φ = 57°. After inishing
of the soil improvement works, the settlements have been observed during the
ongoing construction of roads and bridges. The measured settlements stayed
clearly below the forecasted values (Kirsch et al. 2009).

2.8.3 Vibro stone columns for shipyard

infrastructure in India (2009)
Extensive vibro replacement works have been carried out in Pipavav, India,
to improve the ground for approach roads and hardstand areas of a ship-
building facility, consisting of making facilities for the fabrication of hull
blocks, shiplifting facilities, and multiple land berths.
48 Ground improvement

Soil investigation was carried out in and around the facility by means of
boreholes. Soil proiles for approach roads and hard standing pavements
showed marine clay down to 10 m, with SPT N = values of 2 to 4, underlain
by weathered rock/bedrock with SPT N > 50.
Both the approach roads and hard standing pavements have to be
designed for the heavy trafic loads during the transportation of the 400-
ton shipbuilding hull blocks to the ship assembly area. The long term settle-
ments were to be less than 200 mm and 150 mm for the hard standing
pavements and approach road section, respectively.
Vibro stone columns were chosen as a foundation solution to meet these
settlement limits. The design of the stone columns was carried out in accor-
dance with Priebe’s (1995) method. In order to meet the performance cri-
teria, 900-mm-diameter columns were executed. The average installation
depth was 12 m, with a triangular grid spacing of 2.5 m centre-to-centre. A
section of stone columns along the approach road and hard standing pave-
ment is shown in Figure 2.20.
Several vibro rigs with crane-hung vibrators were used to complete the
installation works. About 144,000 linear metres of vibro replacement col-
umns were installed in 2008 and 2009 to treat a total area of 57,500 m2. To
ensure quality and that the columns were consistently formed, the installa-
tion works were monitored and logged in real time by computers (Raj and
Dikshith 2009).

14 m
road
400 tons
hull block
3m Transport 3 m
median vehicle median
1 m thick pavement
0.00
Moorum filling
–2.5

900 mm ∅
Marine clay stone column

–12.5
Hard strata

Figure 2.20 Typical section of approach road with vibro replacement columns. (From
Vibro replacement columns for shipyard infrastructure at Pipavav, Gujarat,
India. In: Leung C.F., Chu J., Shen R.F. (eds.) Ground Improvement Technologies
and Case Histories, Singapore: Research Publishing Services, pp. 763–769.)
 Deep vibro techniques 49

2.8.4 Vibro concrete columns for a sewage

treatment plant in Malaysia (2008)
Ground improvement technology with VCC was utilised as the foundation
system for the Jelutong Sewage Treatment Plant (JSTP) comprising 12 nos.
of Sequential Batch Reactor (SBR) tanks and associated process tanks.
The subsoil conditions primarily consisted of 5-m-thick reclaimed ill
underlain by about 5 m soft to irm silty clay. Stiff silt was found at a
depth of about 10 m and a dense and hard stratum was encountered at a
depth of more than 40 m. Figure 2.21 shows the extent of waste (domes-
tic garbage) dump, demarcated from trial pits carried out at site, covering
approximately one third of the site. Laboratory results indicated that plas-
ticity index ranges between 20% and 40%; very low sulphate and chloride
content (<0.2%) and average pH value to be around 8, so no additional
protective measures for the cement-based foundation system were required.
The SBR tanks were designed as twin tanks of approximately 90 m × 60 m ×
7 m made out of reinforced concrete.
The foundation system was required to carry SBR tank loads up to 126
kPa with the total settlement of the structure to be less than 75 mm and dif-
ferential settlement to be less than 1:360. VCCs were installed to support
tank and ancillary structures within the garbage area, since soil mixing was
inappropriate. The diameter of each VCC was about 0.6 m with working
loads of 35 tons. Typical spacing of columns was 1.6 m c/c to support a foun-
dation load of 126 kPa. The depth of columns was up to 14 m. The VCCs
were designed to achieve an in-situ unconined compressive strength (UCS)
of around 10 MPa. The columns were installed using custom-built vibro

Pump TNB
Legend: sub-station
station
Vibro concrete columns
N

Vibro stone columns

Cement columns
Waste dump Effluent
disinfection
SBR SBR SBR SBR SBR SBR area
RC
1 2 5 6 9 10
wall
SBR SBR SBR SBR SBR SBR RC wall
3 4 7 8 11 12

Sequential batch reactor tanks

Sludge Gas Anaerobic Sludge
holding storage digestor holding
tank tank tank 800 m

Figure 2.21 Different ground improvement methods for the sewage treatment plant
(From Keller (2009). Brochure 10-65E: Foundation Works for a Sewage
Treatment Plant Using Ground Improvement Methods in Malaysia.)
50 Ground improvement

replacement machines where no water jetting was required. This method

successfully installed the columns without removal of the existing garbage.
To ensure quality, the installation works were monitored in real time by com-
puters to ensure that the columns were consistently formed. Selected VCC
were excavated for examination. It was proven that domestic waste material
was displaced sideways during installation of VCC and did not contaminate
the concrete. Some columns were tested up to 1.5 times the working load
using plate load tests. A total 13 nos. VCC single column and 7 nos. VCC
four-column group load tests were carried out to prove the performance of
the constructed columns. All the load tests were successfully carried out.
Concrete samples for VCC were collected for unconined compression tests.
The retrieved samples were subjected to UCS. Results of tests on VCC sam-
ples showed UCS in the acceptable range of 10–40 MPa (Yee et al. 2008).

2.9 CONCLUSIONS

Vibro systems have proven over 70 years to offer safe and economic meth-
ods of improving weak soils for a wide range of applications. Vibro com-
paction has been used to densify granular soils to signiicant depths (65 m),
and the ability of this technique to reduce the risk of liquefaction during
an earthquake is well documented. Vibro replacement is a widely accepted
technique for improving cohesive and ine-grained soil to support a wide
range of structures. Vibro concrete columns are a good alternative to piles.

REFERENCES

Achmus, M., Wehr, W., and Spannhoff, T. (2010). Building vibrations due to deep
vibro processes. Proceedings of the 7th International Conference on Ground
Improvement Techniques, June 23–25, 2010, Seoul, Korea, pp. 33–42.
Almeida, M.S.S, Jamiolkowski, M., and Peterson, R.W. (1991). Preliminary result
of CPT tests in calcareous Quiou sand. Proceedings of the International
Symposium on Calibration Chamber Testing, June 28–29, 1991, Potsdam,
New York, United States, pp. 41–53.
Arbeitsgruppe Untergrund-Unterbau der Forschungsgesellschaft für das
Straßenwesen (AUFS) (1979). Köln, Merkblatt für die Bodenverdichtung im
Straßenbau.
Baez, J.I. (1995). A Design Model for the Reduction of Soil Liquefaction by Vibro
Stone columns, PhD Thesis, University of Southern California, United States.
Baez, J.I. and Martin, G.R. (1992). Liquefaction observations during installation
of stone columns using the vibro replacement technique, Geotechnical News,
10(3): 41–44.
Balaam, N.P. and Booker J.R. (1981). Analysis of rafts and granular piles,
International Journal for Numerical and Analytical Methods in Geomechanics,
5: 379–403.
 Deep vibro techniques 51

Bell, A.L. (1915). The lateral pressure and resistance of clay and the supporting
power of clay foundations. In: A Century of Soil Mechanics, London, England:
Institution Civil Engineers, pp. 93–134.
Bellotti, R. and Jamiolkowski, M. (1991). Evaluation of CPT and DMT in Crushable
and Silty Sands: Third Interim Report ENEL/CRIS, Milano, Italy.
Bergado, D.T., Chai, J.C., Alfaro, M.C., and Balasubramaniam, A.S. (1994).
Improvement Techniques of Soft Ground in Subsiding and Lowland
Environment. Rotterdam, Netherlands: Balkema.
Brauns, J. (1978). Die Anfangstraglast von Schottersäulen in bindigem Untergrund.
Bautechnik 55(8):263–271.
Brauns, J. (1980). Untergrundverbesserungen mittels Sandpfählen oder
Schottersäulen. Tiefbäu, Ingenieurbau, Straßenbau 8:678–683.
Brennan, A.J. and Madabhushi, S.P.G. (2002). Liquefaction remediation by verti-
cal drain groups, Proceedings of the International Conference on Physical
Modelling in Geotechnics, St. John’s, Newfoundland, July, pp. 533–538.
British Research Establishment. (2000). Speciiying Vibro Stone Columns.
Chambosse, G. and Kirsch, K. (1995). Beitrag zum Entwicklungsstand der
Baugrundverbesserung, Schriftenreihe des Lehrstuhls und Prüfamtes für
Grundbau, Bodenmechanik und Felsmechanik der Technischen Universität
München, Munich, Germany, pp. 411–426.
Covil, C.S., Luk, M.C.W., and Pickles, A.R. (1997). Case history: ground treatment
of the sand ill at the new airport at Chek Lap Kok, Hong Kong. Proceedings
of the 3rd International.Conference on Ground Improvement Geosystems,
London, England, pp. 149–156.
Cudmani, R.O. (2001). Statische, alternierende und dynamische Penetration in nich-
tbindigen Böden, Dissertation, Karlsruhe University.
Cudmani, R.O., Osinov, V.A., Bühler, M.M., and Gudehus, G. (2003). A model
for the evaluation of liquefaction susceptibility in layered soils due to earth-
quakes, 12th Pan-American Conference on Soil Mechanics and Geotechnical
Engineering, Cambridge, Massachusetts, United States, Vol. II, pp. 969–976.
Engelhardt, K.and Golding, H.C. (1975). Field testing to evaluate stone column per-
formance in a seismic area. Géotechnique 25(1):61–69.
European Standard EN 14731. (2005). Ground Treatment by Deep Vibration.
European Standard EN 15237. (2007). Ground Treatment by Vertical Drainage.
Fellin, W. (2000). Advances in Geotechnical Engineering and Tunnelling:
Rütteldruckverdichtung als plastodynamisches Problem, Vol. 3, Rotterdam,
Netherlands: A.A. Balkema.
Foray, P.Y., Nauroy, J.-F., and Colliat, J.L. (1999). Mechanisms governing the
behaviour of carbonate sands and inluence on the design of deep founda-
tions, Proceedings of the 2nd International Conference on Engineering for
Calcareous Sediments Vol. 1, February 21–24, 1999, Bahrain, pp. 55–68.
Forschungsgesellschaft für das Straßenwesen (FGFS). (1979). Merkblatt für die
Untergrundverbesserung durch Tiefenrüttler.
German Patent: Nr. 22 GO 473.
Goughnour, R.R. and Bayuk, A.A. (1979). Analysis of stone column–soil matrix
interaction under vertical load. Paris Colloque International Renforcement
des Sols: Terre Armée et Autres Méthodes, Vol. 1, March 20–22, 1979, Paris,
France, pp. 271–277.
52 Ground improvement

Greenwood, D.A. (1976). Discussion. In: Ground Treatment by Deep Compaction.

London, England: Thomas Telford for the Institution of Civil Engineers, p. 123.
Greenwood, D.A. and Kirsch, K. (1983). Specialist ground treatment by vibra-
tory and dynamic methods. Proceedings of the International Conference on
Advances in Piling and Ground Treatment for Foundations, March 2–4, 1983,
London, England, pp. 17–45.
Haß, H., Wernecke, R., and Kessler, S. (2010). Tiefenverdichtung künstlicher
Inseln in Dubai, Vorträgen auf der 31st Baugrundtagung mit Fachausstellung
Geotechnik, Technischen Universität München, November 3–5, 2010, Munich,
Germany, pp. 185–190.
Herle I., Wehr. W., and Arnold M. (2008). Soil improvement with vibrated stone
columns – inluence of pressure level and relative density on friction angle,
Proceedings of the 2nd International Workshop on Geotechnics of Soft
Soils – Focus on Ground Improvement, September 3–5, 2008, University of
Strathclyde, Glasgow, Scotland, pp. 235–240.
Hu, W. (1995). Physical Modelling of Group Behaviour of Stone Column
Foundations, PhD Thesis, University of Glasgow.
Institution of Civil Engineers. (1987). Speciication for Ground Treatment. London:
England: Thomas Telford.
Japanese Geotechnical Society. (1998). Remedial Measures against Soil Liquefaction,
from Investigation and Design to Implementation. Rotterdam, Netherlands: A.
A. Balkema.
Jebe, W. and Bartels, K. (1983). The development of compaction methods with vibra-
tors from 1976 to 1982, Proceedings of the 8th European Conference on Soil
Mechanics and Foundation Engineering organized by the Finnish Geotechnical
Society, Helsinki, May 23–26, 1983, pp. 259–266.
Keller (1993). Brochure 13-21D: Sanierung des Hochwasserdammes der Leitha bei
Rohrau Pachfurt in Niederösterreich.
Keller (1997). Brochure: Rütteldruckverdichtung, Hamburg Elbtunnel, Vierte Röhre.
Keller (2002). Brochure 11-31E: Offshore Vibro, Compaction for Breakwater
Construction, Seabird Project, Karwar, India.
Keller (2009). Brochure 10-65E: Foundation Works for a Sewage Treatment Plant
Using Ground Improvement Methods in Malaysia.
Kempfert, H.-G., and Stadel, M. (1995). Zum Tragverhalten geokunststoffbewehrter
Erdbauwerke über pfahlähnlichen Traggliedern. Geotechnik Sonderheft zur 4:
Informations und Vortragsveranstaltung über Kunststoffe in der Geotechnik
München, Verlag Deutsche Gesellschaft für Geotechnik, March 16–17, 1995,
Essen, Germany, pp. 146–152.
Kirsch, F. (2004). Experimentelle und numerische Untersuchungen zum Tragverhalten
von Rüttelstopfsäulengruppen, Mitteilung des Instituts für Grundbau und
Bodenmechanik, PhD Dissertation, Technische Universität Braunschweig.
Kirsch, F., Breitsprecher, G., and Rawitzer, S. (2009). Flughafen Berlin – Brandenburg
International – Dimensionierung und Ausführung einer Baugrundverbesserung
für Bauwerke und Verkehrslächen der landseitigen Anbindung,’
Veröffentlichungen des Grundbauinstitutes der Technischen Universität Berlin,
Hans Lorenz Symposium, January 10, 2009, pp. 105–116.
Kirsch, K. (1979). Erfahrungen mit der Baugrundverbesserung durch Tiefenrüttler.
Geotechnik 1:21–32.
 Deep vibro techniques 53

Kirsch, K. (1993). Baugrundverbesserung mit Tiefenrüttlern. 40 Jahre Spezialtiefbau:

1953–1993, Düsseldorf, Germany: Technische und rechtliche Entwicklungen.
Kirsch, K. and Chambosse, G. (1981). Deep vibratory compaction provides founda-
tions for two major overseas projects. Ground Engineering 14(8): 31–35.
Kirsch, K. and Kirsch, F. (2010). Ground Improvement by Deep Vibratory Methods,
New York: Spon Press.
Massarsch, K.R. (1991). Deep Soil Compaction Using Vibratory Probes in Deep
Foundation Improvement, STP 1089, ASTM.
Massarsch, K.R. (1994). Design aspects of deep soil compaction, Proceedings of
Seminar on Ground Improvement Methods, Geotechnical Division, Hong
Kong Institution of Engineers, May 19, 1994, pp. 61–74.
Meier, T. (2009). Application of Hypoplastic and Viscohypoplastic Constitutive
Models for Geotechnical Problems, PhD Thesis, Veröffentlichungen des
Institutes für Bodenmechanik und Felsmechanik der Universität Fridericiana
in Karlsruhe, Heft 171.
Mitchell, J.K. and Huber, T.R. (1983). Stone columns foundation for a wastewater
treatment plant—a case history. Geotechnical Engineering 14:165–185.
Mitchell, J.K. and Solmayr, Z.V. (1984). Time dependent strength gain in freshly
deposited or densiied sand, Journal of ASCE, GT11, 110:1559–1576.
Mitchell, J.K. and Wentz, J.R. (1991). Performance of Improved Ground dur-
ing Loma Prieta Earthquake. Report No. UCB/EERC-91/12, Earthquake
Engineering Research Center, University of California, Berkeley.
Moseley, M.P. and Priebe, H.J. (1993). Vibro techniques. In: Moseley, M.P. (ed.)
Ground Improvement, London, England: Blackie Academic & Professional.
Perlea, V. (2000). Liquefaction of cohesive soils, Soil Dynamics and Liquefaction
2000: Proceedings of Sessions of Geo-Denver 2000, August 5–8, 2000, ASCE
Geotechnical Special Publication No. 107, pp. 58–76.
Priebe, H.J. (1995). The design of vibro replacement, Ground Engineering, December,
pp. 31–37.
Priebe, H. (1998). Vibro replacement to prevent earthquake induced liquefaction,
Ground Engineering, September, pp. 30–33.
Priebe, H.J. (2003). Zur Bemessung von Rüttelstopfverdichtungen – Anwendung des
Verfahrens bei extrem weichen Böden, Bautechnik, pp. 380–384.
Raithel, M. and Kempfert, H.-G. (2000). Calculation models for dam founda-
tions with geotextile coated sand columns, Proceedings of GeoEng 2000,
an International Conference on Geotechnical & Geological Engineering,
November 19–24, 2000, Melbourne, Australia.
Raj, D. and Dikshith, C.V. (2009). Vibro replacement columns for shipyard infra-
structure at Pipavav, Gujarat, India. In: Leung C.F., Chu J., Shen R.F. (eds.)
Ground Improvement Technologies and Case Histories, Singapore: Research
Publishing Services, pp. 763–769.
Raju, V.R. (2002). Vibro replacement for high earth embankments and bridge abut-
ment slopes in Putrajaya, Malaysia, Proceedings for the 4th International
Conference on Ground Improvement Techniques, March 26–28, 2002, Kuala
Lumpur, Malaysia, pp. 607–614.
Raju, V.R. and Hoffmann, G. (1996). Treatment of tin mine tailings in Kuala Lumpur
using vibro replacement. Proceedings of the 12th Southeast Asian Geotechnical
Conference, May 6–10, 1996, Kuala Lumpur, Malaysia.
54 Ground improvement

Schmertmann, J.H. (1991). The mechanical aging of soils. Journal of Geotechnical

Engineering 117(9):1288–1330.
Schneider, H. (1938). Das Rütteldruckverfahren und seine Anwendungen im Erd-
und Betonbau, Beton und Eisen 37(1).
Schüßler, M. (2002). Anwendung neuer, innovativer Gründungslösungen- ist
das Risiko für den Auftraggeber überschaubar? Deutsche Gesellschaft für
Geotechnik Vorträge der Baugrundtagung, September 25–28, 2002, Mainz,
Germany, pp. 339–345.
Schweiger, H.F. (1990). Finite Element Berechnung von Rüttelstopfverdichtungen.
Vorträge des 5. Christian Veder Kolloquium, Technischen Universität Graz,
April 26-27, 1990, Graz, Austria, pp. 1–15.
Seed, H.B. and Booker, J.R. (1976). Stabilisation of Potentially Liqueiable Sand
Deposits Using Gravel Drain Systems, Report No. EERC76-10, University of
California–Berkeley, United States.
Seed, H.B., Idriss, I.M., and Arango, I. (1983). Evaluation of liquefaction poten-
tial using ield performance data, ASCE, Journal of Geotechnical Engineering
109(3): 458–483.
Sidak, N., Strauch, G., and Wehr, J. (2004). Installation of geotextile covered stone
columns by the Keller depth vibrator techniques, Proceedings of ASEP-GI:
International Symposium on Ground Improvement, September 9–10, 2004,
Paris, France, pp. 269–284.
Slocombe, B.C., Bell, A.L., and May, R.E. (2000). The in-situ densiication of granular
inill within two cofferdams for seismic resistance. Workshop on Compaction of
Soils, Granulates and Powders, February 28–29, 2000, Innsbruck, Austria, pp. 33.
Slocombe, B.C, and Moseley, M.P. (1991). The Testing and Instrumentation of Stone
Columns, ASTM STP 1089.
Smoltczyk, U. and Hilmer, K. (1994). Baugrundverbesserung. Grundbau-
Taschenbuch, 5th ed., vol. 2.
Sondermann, W. and Jebe, W. (1996). Methoden zur Baugrundverbesserung für den
Neu- und Ausbau von Bahnstrecken auf Hochgeschwindigkeitslinien. Deutsche
Gesellschaft für Geotechnik Vorträge der Baugrundtagung, September 25–27,
1996, Berlin, Germany, pp. 259–279.
Soydemir, C., Swekowsky F., Baez, J.I., and Mooney, J. (1997). Ground improve-
ment at Albany Airport. In: Schaefer V.R. (ed.) Ground Improvement, Ground
Reinforcement, Ground Treatment: Developments 1987–1997, Geotechnical
Special Publication No. 69, Logan, UT: American Society of Civil Engineers.
Soyez, B. (1987). Bemessung von Stopfverdichtungen. BMT. April, pp. 170–185.
Tomlinson, M.J. (1980). Foundation Design and Construction, New York: Pitman
Publishing.
Topolnicki, M. (1996). Case history of a geogrid-reinforced embankment supported
on vibro concrete columns. In: deGroot, M.B., den Hoedt, G., Termaat, R.J.
(eds.) Geosynthetics: Applications, Design and Construction, Rotterdam,
Netherlands: Balkema, pp. 333–340.
Trunk, U., Heerten, G., Paul, A., and Reuter, E. (2004). Geogrid wrapped vibro
stone columns, Proceedings of the Eurogeo 3, 3rd European Geosynthetics
Conference, March 1–3, 2004, Munich, Germany, pp. 289–294.
US Department of Transportation. (1983). Design and Construction of Stone
Columns, Vol. 1.
 Deep vibro techniques 55

Van Impe, W.F. and Madhav, M.R. (1992). Analysis and settlement of dilating stone
column reinforced soil. ÖIAZ: Österreichische Ingenieur- und Architekten-
Zeitschrift 137(3):114–121.
Vesic, A.S. (1965). Ultimate loads and settlements of deep foundations in sand.
Proceedings of the Symposium on Bearing Capacity and Settlement of
Foundations in Sand, Duke University, Durham, North Carolina, United
States, pp. 53–68.
Wehr, J. and Raju, V.R. (2002). On- and offshore vibro compaction for a crude
oil pipeline on Jurong Island, Singapore; International Conference on Ground
Improvement Techniques, Malaysia, pp. 731–736.
Wehr, W. (1999). Schottersäulen – das Verhalten von einzelnen Säulen und
Säulengruppen, Geotechnik 22(1):40–47.
Wehr, W. (2004). Stone columns—single columns and group behaviour, Proceedings
of the 5th International Conference on Ground Improvement Techniques,
March 22–23, 2004, Kuala Lumpur, Malaysia, pp. 329–340.
Wehr, W. (2005). Variation der Frequenz von Tiefenrüttlern zur Optimierung der
Rütteldruckverdichtung, Proceedings of the 1st Hans-Lorenz-Symposium,
October 13, 2005, Berlin, Germany, pp. 67–78.
Wehr, W. (2005a). Inluence of the carbonate content of sand on vibro compaction,
Proceedings of the 6th International Conference on Ground Improvement
Techniques, July 18–19, 2005, Coimbra, Portugal, pp. 625–632.
Wehr, W. (2006). The undrained cohesion of the soil as criterion for the column instal-
lation effect with a depth vibrator, Transvib 2006: International Symposium
on Vibratory Pile Driving & Deep Soil Vibratory Compaction, November 12,
2005, Paris, France, pp. 157–162.
Wehr, W. (2006a). Stone columns – group behaviour and inluence of footing lexibil-
ity. Proceedings of the Sixth European Conference on Numerical Methods in
Geotechnical Engineering, September 6–8, 2005, Graz, Austria, pp. 767–772.
Wehr, W. (2008). Offshore Schottersäulen in Singapur und Malaysia, Darmstädter
Geotechnik Kolloquium, Technischen Universität Darmstädt, March 13, 2008,
Darmstädt, Germany, pp. 85–93.
Yee, Y.W., Chua, C.G., and Yandamuri, H.K. (2009). Foundation works for a sewage
treatment plant using ground improvement methods in Malaysia. In: Leung
C.F., Chu J., Shen R.F. (eds.) Ground Improvement Technologies and Case
Histories, Singapore: Research Publishing Services, pp. 677–684.
Zöhrer, A., Wehr, W., and Stelte M. (2010). Is ground engineering environmentally
friendly? Proceedings of the 11th International EFFC-DFI Conference, Session
3: Sustainability in the Foundation Industry, May 26–28, 2010, London,
England, on DVD.
Chapter 3

Dynamic compaction
Barry Slocombe

CONTENTS

3.1 Introduction ................................................................................... 57

3.2 History ........................................................................................... 58
3.3 Plant and equipment ....................................................................... 58
3.4 Terminology ................................................................................... 60
3.5 How dynamic compaction works ................................................... 61
3.5.1 Granular soils ...................................................................... 62
3.5.2 Cohesive soils....................................................................... 65
3.5.3 Landills............................................................................... 68
3.5.4 Collapsible soils ................................................................... 69
3.6 Site investigation............................................................................. 70
3.7 Depth of treatment ......................................................................... 70
3.8 Environmental considerations ........................................................ 72
3.9 Practical aspects ............................................................................. 74
3.10 Induced settlement.......................................................................... 75
3.11 Additional comments...................................................................... 76
3.12 Testing ............................................................................................ 78
3.12.1 Standard penetration test..................................................... 79
3.12.2 Pressuremeter....................................................................... 80
3.12.3 Dynamic cone test ............................................................... 80
3.12.4 Static cone penetration test .................................................. 80
3.12.5 Dilatometer.......................................................................... 81
3.13 Case histories.................................................................................. 81
3.14 Concluding remarks ....................................................................... 82
Acknowledgment .................................................................................... 84
References ............................................................................................... 84

3.1 INTRODUCTION

Dynamic compaction (DC) improves weak soils by controlled high-energy

tamping where the applied energies can be greater than 100 times that of

57
58 Ground improvement

driven piling. The reaction of soils during dynamic compaction treatment

varies with soil type and energy input. A comprehensive understanding of
soil behaviour, combined with experience of the technique, is therefore vital
to successful improvement of the ground. Dynamic compaction is capable
of achieving signiicant improvement to substantial depth, often with con-
siderable economy when compared to other geotechnical solutions.

3.2 HISTORY

The principle of dropping heavy weights on the ground surface to improve

soils at depth has attracted many claims for its earliest use. Early Chinese
drawings suggested the technique could be several centuries old (Menard
and Broise, 1976). Kerisel (1985) reports that the Romans used it for con-
struction, and Lundwall (1968) reports that an old war cannon was used
to compact ground in 1871. In the twentieth century, compaction was pro-
vided to an airport in China and a port area in Dublin during the 1940s
and to an oil tank in South Africa in 1955. However, the advent of large
crawler cranes led to the current high-energy tamping levels irst being per-
formed on a regular basis in France in 1970 and subsequently in Britain in
1973 and in North America in 1975.
An extension of the concept of weights dropped onto the ground, rapid
impact compaction (RIC), was developed in England in the late 1970s for
the rapid repair of explosion damage to military airield runways using
modiied (BSP) hydraulic piling hammers acting on a steel foot that remains
in contact with the ground.
A further extension is for large three- and ive-sided towed rollers, also
called impact roller compaction, to compact ground. These and the RIC
equipment apply energy ‘from the top down’ to limited depths that are eas-
ily achieved using lower than normal DC drop heights.

3.3 PLANT AND EQUIPMENT

At irst sight, the physical performance of dynamic compaction would

appear to be simple, using a crane of suficient capacity to drop a suitable
size of weight in virtual free-fall from a certain drop height. Most contracts
are performed with standard crawler cranes, albeit slightly modiied for
safety reasons and productivity, with a single lifting rope attached to the top
of the weight (Figure 3.1). Details such as crane counterbalance weights, jib
lexure, torque convertors, line pull, drum size, type and diameter of ropes,
clutch, brakes, as well as many other factors and methods of working have
been subjected to rigorous analysis by the major specialist organisations
to improve reliability and productivity. The operation must be performed
 Dynamic compaction 59

Figure 3.1 Typical crawler crane and equipment.

safely. As a result the Health and Safety Executive in Britain requires that a
crane should operate at not more than 80% of its safe working load. Some
cranes are better suited than others to the rigours of this type of work, even
though on paper they appear to be of similar capacity.
Recent crane developments allow automation of the whole work cycle.
This is controlled by a data processing unit that plots for each compaction
point its location, number, weight size, drop height, number of blows, and
measurement of imprint achieved. A particular feature of one European
crane is the free-fall winch, which adjusts the rope length automatically
after each blow. Some cranes include the ability for synchronous operation
of two winches to lift larger weights than the conventional crane rating.
The majority of British and American contracts have utilised weights
within the range of 6 to 20 tonnes dropped from heights of up to 20 m.
The majority of UK work is now performed using 8-tonne weights dropped
from heights of up to 12 m. Standard crawler cranes have also been used in
America for weights of up to 33 tonnes and 30 m height. Specialist lifting
frames with quick release mechanisms have been utilised to drop weights
of up to 50 tonnes, and Menard built equipment to drop 170 tonnes from
22 m height in France. In America, and increasingly in Britain, the system
is known as dynamic deep compaction.
Weights are typically constructed using toughened steel plate, box-steel
and concrete, or suitably reinforced mass concrete where durability is the
prime requirement. The effect of different sizes and shapes of the weight
has also been extensively researched with narrower weights generally
being used to speciically drive material down to depth to form dynamic
60 Ground improvement

replacement columns in peaty or Sabkha soils. Treatment has also been

performed below water using barge-mounted cranes and more streamlined
weights with holes cut out to reduce water resistance and increase impact
velocity on the sea bed.
Within the UK RIC typically employs a 7-tonne weight dropped repeat-
edly through a 1.2 m height onto a 1.5 m diameter steel articulated
compaction foot. Whilst the energy per blow is not large (typically 8.4
tonne-metres), the equipment permits a large number of impacts to be
applied at a rate of about 40 blows per minute for typical treatment depth
of up to 3.0 m. Weights from 5 to 12 tonnes are used worldwide. It is how-
ever less successful at treating the mixed soils generally encountered in the
United Kingdom.

3.4 TERMINOLOGY

The original concept for dynamic compaction was to collapse voids, particu-
larly for the treatment of natural sands plus granular, mixed, and cohesive
ills. This was then extended to iner natural soils where the high impact
energy effectively provided localised surcharge to squeeze water out of silts
and clays, this being termed dynamic consolidation. Dynamic replacement
was then developed to drive large-diameter columns of coarse imported mate-
rials through soft near-surface soils, particularly for peat and Sabkha strata.
The worldwide use of dynamic compaction has resulted in a large number
of important terms, some of which can have different meanings to differ-
ent nationalities or could be confused with other geotechnical descriptions.
The following terms have been adopted in Britain:

(1) ‘Effective depth of treatment’ is the maximum depth at which signii-

cant improvement is measurable. The ‘zone of major improvement’ is
typically 1/2 to 2/3 of this effective depth.
(2) ‘Drop energy’ is the energy per blow (i.e., mass multiplied by the
drop height [tonne-metres]).
(3) ‘Tamping pass’ is the performance of each grid pattern over the
whole treatment area.
(4) ‘Total energy’ is the summation of the energy of each tamping pass
(i.e., number of drops multiplied by the drop energy divided by the
respective grid areas [normally expressed in tonne-metre/m2]). It is
not the summation of the drop energy divided by the plan area of the
DC weight.
(5) ‘Grid area’ is the treatment area per drop location for each individ-
ual treatment pass. The dimension of the irst pass is often approxi-
mately equal to the target depth of treatment. Hence for 8 m target
depth the grid area is about 8 × 8 = 64 m 2 .
 Dynamic compaction 61

(6) ‘Recovery period’ is the time allowed between tamping passes to per-
mit the excess pore pressures to dissipate to a low enough level for
the next pass.
(7) ‘Induced settlement’ is the average reduction in general site levels as
a result of the treatment.
(8) ‘Threshold energy’ is the energy input beyond which no further
improvement can practically be achieved or where adverse response
starts to develop.
(9) ‘Overtamping’ is a condition in which the threshold energy has been
exceeded, sometimes deliberately, causing remoulding and dilation
of the soil.
(10) ‘Shape test’ is the detailed measurement of a single or group of
imprint volumes and surrounding heave or draw-down effect, which
permits comparison of overall volumetric change with increasing
energy input.
(11) ‘Imprint’ is the crater formed by the weight at a tamping location.

3.5 HOW DYNAMIC COMPACTION WORKS

In contrast to having constructed a vibro stone or concrete column, the

treatment at that location then being completed, dynamic compaction
(DC), whether to shallow or deep layers, improves the ground to the basal
layers irst and then progressively up to the upper layers in a series of tamp-
ing passes. In contrast, RIC and the three- and ive-sided rollers improve
the soils by irst creating a ‘plug’ or surface layer of denser ground and then
progressively driving this plug/layer to greater depth. The response of the
ground to these two approaches is fundamentally different. Whereas the
relatively lesser number of high-energy impacts at wide-grid centres of DC
tends to initially bypass the upper layers and then by subsequent progres-
sive treatment builds up the strength of the near-surface soils, the larger
number of lower energy impacts of the RIC and roller require consideration
of the possible generation of pore-water pressures that inhibit the required
ground improvement in iner-grained soils.
There is then a fundamental difference between the responses of granular
and cohesive soils when subjected to the high-energy impacts of the process.
It is normal to visualise treatment as a series of heavy tamping passes with
different combinations of energy levels designed to achieve improvement to
speciic layers within the depth to be treated. The most common approach
is to consider the ground in three layers. The irst tamping pass is aimed at
treating the deepest layer by adopting a relatively wide grid pattern and a
suitable number of drops from the full-height capability of the crane. The
middle layer is then treated by an intermediate grid, often the midpoint
of the irst pass or half the initial grid, with a lesser number of drops and
62 Ground improvement

reduced drop height. The surface layers then receive a continual tamp of
a small number of drops from low height on a continuous pattern. It is
sometimes feasible to combine, and sometimes necessary to subdivide, the
basic tamping passes for the reasons outlined in the subsequent discussion.
The performance of increasing correctly controlled total energy input
will normally lead to better engineering performance of the treated ground.
However, analysis of several hundred contracts where in-situ and large load-
ing tests were performed has shown that this is not a linear relationship and
that the post-treatment parameters are heavily dependent upon the charac-
teristics of the soil. As a general rule, similar total energies, whether per m 2
of area or m3 of treatment depth, provide better performance to granular
than mixed soils. Mixed soils are then better treatable than cohesive, with
refuse-contaminated soils generally offering the least performance.
For treatment using RIC the operator monitors and can record the num-
ber of impacts, the total energy input applied, the foot penetration per blow,
and the cumulative penetration. When a speciied parameter is reached, for
example, foot penetration or set per blow, the RIC equipment is moved
and positioned at the next treatment point. This primary treatment pass is
normally performed on a closely spaced grid pattern, typically 1.5 to 2.5 m.
Additional offset and/or lower energy passes, or conventional proof roller
compaction, are occasionally performed to achieve better coverage.

3.5.1 Granular soils

In dry granular materials (i.e., sand, gravel, ash, brick, rock, slag, etc.), it
is very easy to understand how tamping improves engineering properties.
Physical displacement of particles and, to a lesser extent, low frequency
excitation will reduce void ratio and increase relative density to provide
improved load bearing and enhanced settlement characteristics. A feature
that often develops when providing treatment to coarse ill materials is the
formation of a hard ‘plug’ that inhibits penetration of stress impulses to
the deeper layers but is very useful in providing superior settlement perfor-
mance beneath isolated foundation bases. Dynamic compaction can also
improve ‘dirtier’ sands with higher ines content than the vibro compaction
technique when performed without the addition of stone, see Figure 3.9.
When granular materials extend below the water table, a high proportion
of the dynamic impulse is transferred to the pore water which, after a suit-
able number of surface impacts, eventually rises in pressure to a suficient
level to induce liquefaction. This is the theory irst proposed by Menard
and is a phenomenon very similar to that occurring during earthquakes.
Clearly the existing density and grading of the soils will be major factors
in the speed at which this liqueied state will be achieved. Low frequency
vibrations caused by further stress impulses will then reorganise the par-
ticles into a denser state. This is comparable to the response of sands from
 Dynamic compaction 63

the vibro compaction technique for which D’Appolonia (1953) suggested

that a vibrational acceleration in excess of 0.5 g was necessary to achieve
such a densiication effect.
Dissipation of the pore-water pressures, in conjunction with the effective
surcharge of the liqueied layer by the soils above, results in further increase
in relative density over a relatively short period of time. This can vary from
1 to 2 days for well-graded sand and gravel, to 1 to 2 weeks for sandy silts
and varies with the applied energy. The testing programme should there-
fore recognise the time-dependent response for soils that are normally con-
sidered to be free-draining. Longer-term improvement, possibly as a result
of chemical bonding or high residual lateral stresses within the soil matrix,
has been reported by Mitchell and Solymar (1984).
There is, however, another school of thought in which the aim is to
avoid the liqueied state. While it is recognised that liquefaction cannot
be avoided in deep, loose, sandy deposits with a high water table, as are
often encountered in parts of North America, the Middle and Far East,

Original level
Level after tamping
Hard
plug

Water table

Volume
(positive)
Resultant
(loose soils)
Imprint volume
(loose soils)

Resultant
(dense soils)

Draw down
(loose soils)

Increasing number of drops

Figure 3.2 Volumetric response—granular soils.

64 Ground improvement

Tip resistance (MPa) Friction ratio (%)

0 10 20 4 2 0
0

3
5
83 B

10

Depth (m) Comparison of

GPT 83B (before)

GPT 3 (after first pass)

GPT 6 (after treatment)

Figure 3.3 In-situ test results—granular soils.

such conditions are rare in Britain. The treatment is therefore designed to

provide compaction by displacement without dilation or high excess pore
pressures by using a smaller number of drops from a lower drop height.
This method requires substantially lower energy input than the liquefaction
approach, with consequent economies. Laboratory and in-situ tests have
consistently shown that in order to achieve maximum density, the lowest
number of stress impulses to attain the required energy input will provide
the optimum result. Saturated granular soils will normally require higher
treatment energy overall, in a larger number of tamping passes, than if the
soils were essentially dry.
Figure 3.2 illustrates the typical volumetric response for granular soils
and Figure 3.3 illustrates electric cone results for a site of clean sand with a
water table at about 2.5 m depth treated by 15-tonne equipment. This also
illustrates improvement with time since the second tamping pass was only
capable of treating to about 4 m depth.
When the individual sand particles are weak, such as the calcareous
sands of the Middle East, ‘sugar’ sands of North America, or the Thanet
Sands of Britain, crushing tends to occur during the treatment. A similar
response affects ash, clinker, and weak aerated slags. When these soils are
dry, the effect of such particle breakdown is not particularly signiicant.
 Dynamic compaction 65

However, below the water table the higher proportion of ines developing
with increasing energy input results in a rapid change from a granular to a
pseudo-cohesive soil response.
The existence of very dense layers within the ground can cause anom-
alous results. Where, for example, cemented layers occur within natural
sands, these tend to absorb the energy impulse and arch over the underlying
stratum. A similar phenomenon can occur with vibro treatment where the
cemented zones do not collapse around the vibrator to permit densiica-
tion to occur. In these situations, where they occur at shallow depth, the
dynamic compaction will break up the cemented layer. However, at greater
depth, the energy levels required to break the stratum may be beyond the
capabilities of the equipment on site. The presence of such layers is often
not adequately revealed by normal site investigation.
In summary, excellent engineering performance can easily be achieved in
dry granular soils using both DC and RIC equipment. However, care must
be exercised for the treatment of soils with signiicant silt content, particu-
larly below the water table.

3.5.2 Cohesive soils

The response of clays is more complex than that of granular soils. There is
again the distinction between above and below the water table.
With conventional consolidation theory, a static surcharge loading
will collapse voids within clay ills and expel water to induce consolida-
tion and increase strength. The rate at which this occurs is dependent
upon the imposed load, coeficient of consolidation, and length of drain-
age path. In contrast, dynamic compaction applies a virtually instanta-
neous localised surcharge that collapses voids and transfers energy to the
pore water. This creates zones of positive water pressure gradient that
induce water to drain rapidly from the soils matrix. This effect is further
accelerated by the formation of additional drainage paths by shear and
hydraulic fracture. Consolidation therefore occurs much more rapidly
than would be the case with static loading. Dynamic compaction lit-
erally squeezes water out of the soil to effectively preload the ground.
However, as with staged construction, the application of too high energy
too soon can lead to problems. A typical volumetric response is illus-
trated in Figure 3.4.
Where the soils occur above the water table, the clays tend to be of rela-
tively low moisture content, generally less than their plastic limit, where
even a small reduction in water content can result in signiicant improve-
ment in bearing capacity. As such, treatment is relatively straightforward
and is mainly the collapse of air voids to provide a more intact soil struc-
ture. Care must however be exercised for the treatment of clays that are
wetter than their plastic limit and higher plasticity clays.
66 Ground improvement

Level after tamping

Original level
Shear planes

Water table
Volume
(positive)

Resultant (wet soils)

Imprint volume

Resultant
(dry soils)

Increasing number of drops

Volume Heave around

(negative) single imprint

Figure 3.4 Volumetric response—cohesive soils.

Where the clays occur below the water table, a much larger reduction in
moisture content is generally required in the presence of a smaller available
pore-pressure gradient and a longer drainage path. These conditions can, if
not properly controlled, result in the threshold energy being achieved much
more rapidly and resulting in localised failure of the clay matrix. Control is
then achieved by using greater numbers of tamping passes of lower energy
input, requiring greatly extended contract periods in comparison to normal
productivity.
To date, only nominal degrees of improvement have been achieved in thick
layers of relatively weak saturated alluvial clays and silts, even with addi-
tional measures such as drainage trenches illed with sand or wick drains.
Where such layers are relatively thin and require treatment, a better speed
of response is recorded due to the shorter drainage path. In some instances
coarse granular material is driven into these materials to provide better grad-
ing that is more suited to treatment, or to displace from speciic locations
beneath part of a building area. It is, however, more common in the UK to
adopt vibro stone columns in such soils to more critical locations, such as
more heavily loaded foundations, and then perform dynamic compaction to
 Dynamic compaction 67

preload the ground (see Slocombe 1989) with the beneit of the stiffer col-
umns also acting as drains to control excess pore-water pressures.
For predominantly clay-type ill materials above the water table, the
clay lumps can be considered as large weak particles of almost granular
response. However, the major improvement is achieved by collapsing voids
to provide a more intact structure. Clearly the strength of the lumps and
sensitivity of the clay is of paramount importance in such soils. Differing
degrees of weathering can also give rise to markedly variant responses on
a site and experienced observation is required to deine such locations.
Mudstone and shale fragments can break down to a material of clayey
response, particularly when heavy rainfall occurs.
For clay-type ills below the water table, the voided structure allows
higher mobility of water causing lower excess pore pressures and shorter
recovery periods in comparison to natural clays. The constituents would
be of higher moisture content but again improvement would be achieved
mainly by collapse of voids. Monitoring of excess pore-water pressures by
means of piezometers is clearly useful but problematic above the water table.
It cannot be emphasised too strongly that the treatment of clay ills and
clay soils requires experienced control on site. During treatment, after a
small number of drops, heave starts to develop around the edges of each
imprint. If tamping continues, the heave can build up to such an extent that
it can exceed the volume of the imprint. Clearly this is the precise opposite
of what is desired. Also, additive heave can occur by the performance of the
adjacent tamping position at too narrow a grid dimension.
Particular care has to be exercised in the timing of successive tamping
passes to permit adequate recovery of pore pressures to avoid excessive
remoulding of the soils. Such approach can however be relatively slow and,
in view of the emphasis placed these days on productivity, the vibro stone
column in advance of dynamic compaction method described above is
sometimes adopted.
If excessive heave around an individual imprint does start to occur, it
is essential that the tamping at that position be stopped. This may only
extend over a relatively conined area with better ground elsewhere. In soft
areas it is better that twice the number of lighter energy input tamping
passes be performed in a ‘softly softly’ approach.
Similar considerations apply when attempting to provide treatment to a
signiicant depth where the surface layers are clayey. The strength of the
surface soils can reduce in the short term and time has to be spent improv-
ing a disturbed matrix to reconstitute its original, let alone desired, proper-
ties. This is particularly dificult where thick crusts to, say, 2–3 m depth
of stiff to very stiff clays overlay a granular deposit requiring treatment. In
this situation even higher than normal energies are required to attain the
deeper layers giving rise to even greater potential for virtually destroying
the surface soils.
68 Ground improvement

Modulus of deformation (kPa) Pressure limit (kPa)

15000 10000 5000 0 500 1000 1500

2m

4m

6m

8m

Before treatment

15 days after treatment

150 days after treatment

5 years after treatment

Figure 3.5 In-situ test results—cohesive soils.

The treatment of clayey soils will nearly always require a larger number
of tamping passes when compared to a similar proile of predominantly
granular constituents. Eficient treatment is achieved by attempting to pro-
vide as much improvement as quickly as possible while recognising that
the response of the soils will dictate the speed of the treatment operations.
Clay soils will continue to improve for a signiicant period after treat-
ment as reported by West (1976). Figure 3.5 illustrates further measure-
ments on this site taken ive years after treatment.
In summary, dry cohesive ills respond well to dynamic compaction.
Care must be exercised in the treatment of weak natural clayey soils or clay
ills below the water table. The prior performance of vibro stone columns
to both stiffen the ground and enhance drainage has been successfully com-
bined with dynamic compaction to weak clayey soils.

3.5.3 Landfills
The capability of dynamic compaction to treat every square metre of road
and parking areas is increasingly used in the development of former landill
sites where, depending on their age, the original degradable constituents
 Dynamic compaction 69

have decayed to create extensive voids. It has also been performed to reduce
ground levels to avoid costly removal to specialist tip to permit develop-
ment at the desired site level.
As a general rule, the older the landill, the less the residual presence
of matter susceptible to long-term decay and some older ills, particularly
those of high ash content, have been compacted to also support structures
that would normally be piled. However, the more recent ills generally con-
tain signiicant proportions of organic matter and structures would nor-
mally be piled.
There is as yet little documented proof that dynamically compacted
landill can affect the rate of decay of residual degradable constituents,
although a paper by Sharma and Anirban (2007) clearly records far bet-
ter post-treatment performance at creep rate of 2% per log cycle than for
static surcharge over a monitoring period of about 15 years. As there will
be ongoing decay, when this technique is combined with piled structures,
increasing differential settlements will become apparent with time and a
degree of maintenance may be required at some future time.
The principle of treatment to landills is comparable to the treatment of
mixed clayey ills but with generally higher energy input than for inert ills.
This is to collapse near-surface voids and to ‘‘overcompact’’ the remaining
inert constituents. If a void then starts to develop due to localised long-
term decay, the inert materials will tend to ravel into the void, bulk up, and
spread the void effect rather than have a localised sharp deformity in the
inished surface. Geogrids have also been used for a number of sites where
the movements of heavy goods vehicles were critical to the development
operations.
Many landill sites have clay capping with basal clay liners to avoid down-
ward migration of leachate into an aquifer. The DC drop energy should be
limited to avoid shearing of the basal clay liner and care exercised in the
design of the treatment operations to avoid the surface clays developing into
a quagmire when attempting to apply the higher than for inert ill energy
input to signiicant depth.

3.5.4 Collapsible soils

Rollins and Kim (2010) have reported on the successful treatment of natu-
ral cohesionless and low plasticity collapsible soils using typically higher
compactive energy than would be used for noncollapsible soils. In Britain,
there are many former opencast coal sites where the degree of control in
backilling has resulted in the presence of suficient voidage to also pro-
vide potential for collapse settlement upon wetting. In such situations it is
important when constructing vibro stone columns to ‘seal’ the columns to
prevent the ingress of water into the collapsible ills via the free-draining
stone columns.
70 Ground improvement

Dynamic compaction to these sites reduces the near-surface voids to

inhibit the collapse potential. It also reduces the permeability within the
treatment depth to further inhibit the migration of any water into the
ground. Consideration should however be given to how water is drained
away from the development to avoid a possible plume of water extending
beneath the development where there is greater thickness of collapsible ill
than the treatment depth. Again, higher compactive energy than for non-
collapsible soils would normally be performed.
The use of RIC to treat loess to nominal 3.0 m depth in Kazakhstan has
been reported by Serridge and Synac (2006).

3.6 SITE INVESTIGATION

As with vibro designs, the extent of the site investigation should be appro-
priate to the type of development. For deep ill sites, desk studies to estab-
lish the locations of buried high walls plus age and degree of control of
placement are essential data. Water contents for comparison with liquid
and plastic limits should be performed for clayey constituents. Densities
as revealed by SPT and CPTs plus the basal soils, whether clay, sand, and
gravel or rock are also required. The presence of any overhead wires, bur-
ied services, or nearby structures should also be established.

3.7 DEPTH OF TREATMENT

Menard originally proposed that the effective depth of treatment was related
to the metric energy input expression of (WH)0.5 where W is the weight in
tonnes and H the drop height in metres. This was modiied by a factor of
0.5 by Leonards et al. (1980) for relatively coarse, predominantly granu-
lar soils, and factors of 0.375 to 0.7 by Mitchell and Katti (1981) for two
soil types. The most exhaustive analysis yet published has been provided
by Mayne et al. (1984). The author suggests that the range of treatment
depths varies with initial strength, soil type and energy input as illustrated
in Figure 3.6, as well as the depth to the groundwater table. Figure 3.6 sug-
gests that factors as high as 0.9 could apply for shallow depths of loose soils
and as low as 0.25 for deeper treatment.
There are many factors affecting this dimension, not least of which are
the type and competence of the surface layers, position of the water table,
and numbers of drops at each location. Assessment of in-situ results to
determine such depths also tends to be subjective and will be affected by
the recovery period after treatment. As noted in the previous section, a
solid ‘plug’ of very dense material can form beneath the impact locations to
inhibit the improvement to depth. Weak surface soils and a high water table
 Dynamic compaction 71

Depth of
influence (m) D = √WH D = 0.5 √WH
15

Loose or
10 weak soils

Stiffer or
dense soils
5

5 10 15 20 25 30
√WH

Figure 3.6 Depth of treatment.

can also limit the physical performance of a suficient number of stress

impulses to induce only a minor improvement to the basal layers. However,
knowledge of the depth of any stress impulse is a vital factor in both the
planning of the treatment operations and the potential for transmission of
vibrations as discussed in the next section.
Kinetic energy at the point of impact is clearly a major factor in the depth
of treatment and increasing the drop height will increase velocity. In Britain,
high-speed photography has shown typical impact speeds of about 35 and
50 mph for 8 tonnes from 12 m and 12 tonnes from 15 m heights to achieve
effective depths of treatment of about 5 to 6 m and 6 to 8 m, respectively.
For RIC, the BRE Report No 458 (2003) records depths of treatment
of between about 2.0 and 4.0 m at total energy inputs ranging from 80 to
190 tonne-metre/metre2 for granular ills and silty sands. Greater treatment
depths using higher energy inputs in favourable conditions in Japan, Iran,
and Canada, together with a range of applications, have been reported by
Serridge and Synac (2006).
The shape of dynamic compaction improvement in the ground tends to
be similar to the Boussinesq distribution of stresses for a square founda-
tion. Modiication of energy levels for each tamping pass can be used to
custom-design the treatment scheme to the speciic soils proile and engi-
neering requirements. In contrast, the shape of vibro improvement tends to
increase with depth. In earthquake areas the required density of soil from
the Seed and Idriss (1971) analysis is often better provided at depth by the
vibro technique, which has the added advantage of forming stone columns
to act as drains in the iner soils. Stone columns can also be combined to
reinforce weak cohesive soils at a depth that would be dificult to treat
when using dynamic compaction for surface ill layers (Slocombe 1989).
72 Ground improvement

As noted earlier, high-impact energies can weaken the surface layers, and
the aim is therefore to combine effects to achieve improvement through-
out the whole of the desired treatment zone. For example, on a project
in Saudi Arabia, drop height, numbers of drops, and the treatment grid
were adjusted to provide treatment to three distinct sand layers requir-
ing improvement in a single tamping pass (Dobson and Slocombe 1982).
Clearly, if all structures are founded at depth there is no need for the inal
tamping pass for treatment to the surface layers, provided the grid of the
earlier tamping passes produces overlapping effects at the founding level.

3.8 ENVIRONMENTAL CONSIDERATIONS

Dynamic compaction utilises large, highly visible equipment. The process

creates noise and vibration, both of which must be considered in Britain
under the Control of Pollution Act, 1974. The standards listed in the refer-
ence section provide further details (BS 5228, 2009; BS 7385, 1990; BRE
Digest 403, 1995).
Airborne noise levels are generated by a number of causes. Of these the
point of impact is by far the highest noise level at typically 110 to 120dB
at source. However, its duration only occupies about 0.5% of the lifting
cycle. The considerably lower noise values during lifting and idling when
combined with the impact noise using the LAeq calculation method will
normally meet most environmental limitations at distances of greater than
50 m from the treatment operations. Lower than normal noise limits can
be achieved by working within a speciied zone for only a certain number
of hours during the working day. Large plate glass windows can some-
times act as diaphragms to change the noise characteristics inside a prop-
erty. Echoes, wind direction, and angle of crane exhaust are all factors that
should also be considered.
By far the most important consideration, however, is ground vibration. In
addition to the magnitude of the vibration, the typical frequency of about
5 to 15 Hz is potentially damaging to structures and services, and particu-
larly noticeable to human beings. It is suggested that there are three vibra-
tion levels that will inluence the design of the treatment scheme. Guide
values of resultant peak particle velocity at foundation level for buildings
in good condition are

Structural damage 40 mm/s

Minor architectural damage 15 mm/s
Annoyance to occupants 2.5 mm/s

Lower values must be adopted for buildings in poor condition or envi-

ronmentally sensitive situations such as schools, hospitals, and computer
 Dynamic compaction 73

installations. Certain major computer companies recognise the importance

of the vibration frequency by requiring more onerous limits for frequencies
below 14 Hz than above. It should be noted that some ampliication can
occur as the vibration rises up certain types of structure; for example, that
1.0 mm/s at ground level could be 2.5 mm/s at the third loor. Services and
utilities must be considered on an individual basis depending upon their
age, condition, and importance with values of 15 to 20 mm/s normally
being considered acceptable, except for higher pressure gas mains.
The level of vibration transmitted through the ground is an imprecise sci-
ence because of the variable nature of the characteristics of soils. Field mea-
surements of vibrations at ground level have revealed a number of trends,
which are illustrated in Figure 3.7. The upper dynamic compaction limits
tend to occur in the presence of granular or refuse-type soils and the lower
limit in cohesive strata. A high water table will also tend towards the higher
limit. The upper vibro limit is for vibrators operating at a frequency of 30
Hz and the lower for 50 Hz.
Careful assessment is required where the soil being treated is directly
underlain by relatively dense sand, gravel, or rock which will tend to trans-
mit vibrations to larger than normal distances with comparatively little
attenuation. Pre-existing dense surface or buried layers can have a similar

Peak particle velocity

(mm/sec)
50

40

30

20

10

10 20 30 40 50 60 70
Distance (m)
Large weight
Dynamic
Normal compaction

Vibro

Figure 3.7 Vibrations.

74 Ground improvement

effect of causing the transmission of higher than anticipated vibration lev-

els. The physical performance of the treatment work improves interparticle
contact of the soils and, as such, vibration levels can sometimes increase
towards the end of the treatment operations even though the inal impact
energy levels are substantially lower than those performed for the initial
tamping passes.
When vibrations become a problem, there are three main methods of
reducing their effect. The irst is to simply reduce the height of drop and
compensate by increasing the number of drops per imprint. This reduces
both the impact energy and penetration of the stress impulse that may have
attained an underlying dense stratum. The second method of reduction is
to utilise a smaller weight and the third is to excavate a cut-off trench to
suficient depth to intercept the surface wave. Cut-off trenches have been
found to reduce the transmitted vibration levels by about 50% and are most
effective when located near to the structure or service requiring protection.
As the RIC equipment is based on lower impact energy but greater num-
bers of drops, this method has been employed as close as about 10 to 15 m
from an existing structure that was to be extended.
However, human beings are particularly sensitive at detecting vibrations
and have a psychological reaction in believing that damage is caused even
though the values are far below the well-established damage threshold lev-
els. A thorough public relations exercise can sometimes help to overcome
concern among local residents. Building surveys prior to the commencement
of treatment are often advisable. People are often not aware that vibrations
caused by passing lorries or slamming doors can exceed the levels of minor
architectural damage. Similarly, very cold or hot weather and snow loading
can lead to structural movements that are sometimes incorrectly attributed
to vibrational causes.

3.9 PRACTICAL ASPECTS

There are a number of practical factors that must be taken into account
when performing dynamic compaction contracts. The large crawler crane
must be safely supported by a free-draining working surface, the thickness
of which will depend upon the type of ground being treated. If the surface
1.0 m layer is basically granular, no imported working carpet is generally
required. However, when working from a sandy surface, particularly dur-
ing wet weather, ly-debris has been seen ejected through the air up to 60
m from the point of impact. If work is carried out near roads, railways, or
property, a moveable screen is often used to intercept such ly-debris, albeit
these affect productivity. Alternatively, the programme should contain suf-
icient lexibility to permit treatment to be performed within, say, 50 m of
such features only when the surface conditions permit its safe operation.
 Dynamic compaction 75

As the RIC foot remains in contact with the ground, there tends not to be
ly-debris issues with this equipment.
Where cohesive surface conditions exist, a free-draining granular work-
ing carpet is normally required. The thickness can be as little as 150 mm
for light energy treatment in reasonably competent soils up to 1.5 m when
treating heavily voided refuse ills. When aiming for substantial depth of
treatment, thick working carpets of 1.0 m or more have been found to
inhibit the stress impulse. A more eficient operation, which also provides
greater control of backill quantities, is to start with only 0.5 m thick-
ness and to backill imprints directly, thus preserving the working carpet
for successive tamping passes. In such cases it is useful to compare the
imported quantities to the assessed volume of the directly inilled imprints.
The most commonly adopted approach is to blade these preferably coarse
(up to 200 mm or single brick size) granular materials into the localised
deep imprints using a large dozer. In such cases it is useful to perform a
grid of levels before and after treatment to assess the induced settlements.
Winter working will place more onerous requirements on the adequacy
of the working surface. The general rule is to increase the depth of the gran-
ular working carpet by 25% in comparison to summer thickness. When
working in arid climates, there is often no need for any working surface,
even for clayey soils.
As the performance of dynamic compaction tends to induce increases
in water pressures, a pre-existing groundwater table within about 1.0 to
1.5 m of the working level can inhibit the productivity of the technique. In
such cases bottom-feed vibro stone columns may be the preferred approach.
Safe working is a prime consideration. If more than one rig unit is to
be used, they should be separated by at least 30 m. Similarly, subsequent
operations by the main contractor may have to be delayed until the treat-
ment operations are suficiently remote. Whilst dynamic compaction can
be performed over areas of vibro stone columns it has to be performed
before any adjacent piling operations to avoid possibly damaging the con-
structed piles.

3.10 INDUCED SETTLEMENT

The general densiication and collapse of voids will induce general reduc-
tion in site levels, the induced settlement being dependent on the total
energy input and the manner in which it is applied. Initial shape tests are
performed when the soils are loosest. As such, simple extrapolation of these
results will overestimate the amount of induced settlement. Mayne et al.
(1984), as part of his survey of 124 different sites, reported that the magni-
tude of induced settlement depended on the applied total energy input, also
stating that the thickness of the layer was probably an important factor for
76 Ground improvement

Table 3.1 Approximate induced

settlement as percentage of
treatment depth
Soil type % depth
Natural clays 1–3
Clay ills 3–5
Natural sands 3–10
Granular ills 5–15
Refuse and peat 7–20

six soil types. This analysis does not, however, take into account either the
initial softness/density of the soils or the proportion of total energy applied
by the high velocity initial passes or low velocity inal tamping pass that
numerically is very signiicant in determining total energy input. Also, the
application of too high an energy in clayey soils will result in less than opti-
mum induced settlement occurring in practice.
A convenient simple approach is to adopt approximate percentages of
the target treatment depth for 8-tonne (50 to 100 tonne-metres/m 2) and
12-tonne equipment (100 to 200 tonne-metres/m 2), the total energies with
the 12-tonne energy applying to greater depth of treatment.
Higher percentages can be induced. However, the increase in energy
input will not be linear (e.g., to increase from 10%–15% induced volume in
refuse would require 200%–250% of the normal energy because during the
treatment the material becomes progressively stronger and there is less and
less potential void reduction available). Care has to be exercised to avoid
overtreatment and possible loss in strength in these situations, especially
since refuse tips tend to be capped by clay soils.
Loose materials will obviously settle more than denser soils. As noted
earlier, ash and certain types of slag also tend to break down during treat-
ment to produce induced movements towards the higher value for granular
ills given above (see Table 3.1).

3.11 ADDITIONAL COMMENTS

Dynamic compaction is a highly sustainable technique since it does not

use cement or quarried stone, normally only requiring suitable inert free-
draining granular waste as a working platform and to inill localised deep
imprints. It is also an area treatment technique that permits changes in
foundation layouts and localised loadings, for example, mezzanine support
foundations, anywhere within the treatment area. A number of dynamic
compaction contracts have permitted the rebuilding of developments where
ire destroyed the original by simply performing a number of loading tests
upon the treated ground.
 Dynamic compaction 77

Pulverised fuel ash (PFA) comprises relatively ine single-sized particles

that exhibit pozzolanic properties when properly compacted by conven-
tional methods. When tipped into water, the materials tend to locculate
with little self-compaction occurring with time. Dynamic compaction
has been attempted on a number of settlement lagoons with little success.
However, dry PFA is considered suitable for treatment, albeit with a number
of controls. Similarly, weak chalk tends to crush upon high-energy impact
and, when wet, rapidly loses strength. Extreme caution is recommended
when considering the feasibility of treating this weak rock.
Peaty soils can be treated in many different ways, depending on the
required end result. High energy can be applied to physically displace the
material wholescale from beneath the line of a major road. Discrete col-
umns of sand- to cobble-sized ill can be driven into the peat in a manner
similar to stone column theory or normal treatment methods applied to
simply squeeze out some of the water and preload the ground. The basic
fact that must be considered throughout is that peat tends to be a very weak
material of high moisture content. As such, the pre-loading method will
take time to perform.
Dynamic compaction has been performed to collapse shallow solution
cavities. It is important that the extent of these voids be accurately deter-
mined prior to treatment so that the crane is positioned suficiently remote
when the cavity caves in to avoid falling into the void. Considerable care
has to be exercised during the dropping of the weight to avoid the rope pull-
ing off the crane drum or the weight punching into the void and becoming
trapped beneath the surface. Long, narrow weights are better suited to this
operation.
Many sites of former heavy industry are now being reclaimed in Britain.
These often present the designer with the problems of deep ill and massive
obstructions from old basements and foundations. Any technique that has
to make a hole in the ground will experience dificulty in gaining adequate
penetration, and large excavations often have to be performed to remove
the obstructions. However, with dynamic compaction such features can
be left in place provided they occur at suficient depth to avoid excessive
differential performance. For most industrial or low-rise housing develop-
ments, this depth would be a minimum of 1.0 m below the underside of
new foundations or loor slabs, sometimes a depth equal to the width of
the buried foundation. The sequence of operation would be for advance
earthworks to remove all known features down to a speciic level, then
perform treatment and normal construction operations. In choosing this
excavation level, the designer should take into account the type of structure
and its tolerance to some degree of differential performance and the fact
that the treatment will induce a reduction in site level that is slightly higher
than normal as a result of the loose nature of the surface materials after
the pre-excavation operations. It is also normal to apply a slightly higher
78 Ground improvement

than normal energy input to reduce the differential performance between

the areas of massive foundations and abutting weaker soils which, if weak
clay ills, could be signiicant. Where the backill comprises large concrete
posts, slabs, or waste that could arch over a void, care must be taken in the
design of the foundations since the treatment may not necessarily cause the
cavity to wholly collapse and could weaken the member forming the roof to
the void. In these circumstances, higher than normal energies are preferred
to break down such potential.
Brownield sites with minor contamination are well suited to treatment
since the technique does not create a bore that could permit the migra-
tion of leachate. Care should be exercised to ensure that the impact ener-
gies do not shear any underlying clay layer or basal liner to former landill
that may prevent the downward migration of contaminants into an aqui-
fer. Similarly, if there is a high water level that is contaminated, attention
should be paid to the possibility of a rise in level during the treatment and
its effect on adjacent property or features. In some cases installation of
drains lined with HDPE membrane or monitoring may be required.
Areas contaminated with chemicals or asbestos now have to be devel-
oped. The major advantage that dynamic compaction has over alternative
methods is that it can be controlled by inilling deep imprints before they
penetrate through the working carpet into the contaminated soils, to avoid
exposure of hazardous material to the atmosphere while still compacting
soils at depth.
Sites of former quarries are prime situations for treatment in view of
the potential for piles to glance off the buried subvertical face between ill
and rock, and the possibility of constructing piles to inadequate depth as a
result of false readings from boulders or inaccurate historical information
on the depth of the quarry.
Similarly, former opencast coal mining sites are now being treated, even
when the ills are placed to better than 95% maximum dry density, to fur-
ther reduce voids for better settlement control. One such site involving load
testing recorded typical pre-treatment short-term moduli within the range
of 16–18 MPa being increased to 30–40 MPa using 8-tonne weights and
50–60 MPa with 12-tonne weights.

3.12 TESTING

Many contracts have simply involved the measurement of depth of irst

pass imprints and monitoring of site levels. Post-treatment in-situ (SPT
and CPT) and loading tests are often performed, and since the technique
provides treatment to large areas very quickly, the speed at which such
tests provide the necessary information is important, particularly if test-
ing between tamping passes. It is rare, therefore, to recover samples for
 Dynamic compaction 79

Table 3.2 Suitability for testing dynamic compaction

Test Granular Cohesive Comments
Dynamic cone ** * Too insensitive to reveal soil type. Has
dificulty penetrating densely compacted
ground.
Electric cone *** * Particle size important. Can be affected by
lateral earth pressures generated by
treatment. Best test for seismic liquefaction
evaluation in sands.
Boreholes and SPT *** ** Eficiency of test important. Recovers
samples.
Small plate * * Poor coninement to zone being tested.
Affected by pore-water pressures.
Large plate ** * Better conining action.
Skip ** ** Can maintain for extended period.
Zone loading **** **** Best test for realistic comparison with
foundations.
Full-scale ***** ***** Rare.
Key: * least suitable, *****most suitable

laboratory testing. In clayey soils, as with the performance of the treat-

ment, it is essential that suficient recovery period be allowed to avoid
ambiguous results.
It is common for dynamic compaction to be performed for sites under-
lain by coarse ills or including obstructions that would cause penetration
problems for vibro or piling methods. Similar problems could therefore
be reasonably anticipated in attempting to perform in-situ tests. Air drills
have been used on a small number of contracts to predrill a test location to
below the level of the potential obstruction. However, surface loading tests
only are more normally performed in this situation.
Table 3.2 describes the relative merits of various test methods.
Additional comments on the advantages and limitations of certain in-situ
tests follow.

3.12.1 Standard penetration test

This is probably the most useful in-situ test as it is applicable to both granu-
lar and cohesive soils. However, being of a dynamic nature it is particularly
sensitive to the presence of residual pore-water pressures, quickly liquefy-
ing the stratum being tested and producing lower than expected results.
A sample is normally recovered and the speed of provision of information
is adequate for most contracts. The main drawback is that a considerable
amount of time and money can be spent chiselling to penetrate the very
dense surface layers normally provided by the treatment.
80 Ground improvement

20 40 60 80 100 120
Load (tonnes)

10

20

30 Granular

Cohesive
40
Settlement
(mm)

Figure 3.8 Post-treatment zone loading test results.

Tests normally reveal superior performance being achieved when treat-

ing granular materials in comparison to clayey soils; see Figure 3.8 for iden-
tical total energy input and zone test size for different site areas of granular
and cohesive ills.

3.12.2 Pressuremeter
The dynamic compaction technique has been historically associated with
this test. While this method is often used in mainland Europe, it is now
rarely used in Britain.

3.12.3 Dynamic cone test

This is relatively cheap and robust but is limited by the inability to deter-
mine, without the performance of alternative parallel testing, whether
a zone of low blow count is caused by loose zones or cohesive layers
that would be expected to respond differently to the treatment. Since the
majority of ground improvement contracts performed in Britain require
treatment to variable ill sites this method has been found to be of limited
value.

3.12.4 Static cone penetration test

These tests are ideally suited to the testing of sands because they illustrate
the soil type by means of the friction ratio. They are considered less success-
ful in clayey soils since experience has shown that this test is particularly
affected by the presence of residual pore-water pressures. Being relatively
sophisticated it is not recommended that these tests be used in the presence
of coarse ill materials.
 Dynamic compaction 81

3.12.5 Dilatometer
This method would appear to have potential. However, no information has yet
been published for its evaluation of the treatment of the soils and ills of Britain.

3.13 CASE HISTORIES

As part of a larger vibro compaction contract to densify clean sand ills to

up to 18 m depth, dynamic compaction was performed to areas of surface
sands to depths of up to 4 m where vibro was not performed. The sand ills
in the DC areas were by design of signiicantly higher ine contents than the
vibro areas with up to 40% total ines content, typical CPT friction ratios
being up to 2.0% plus local pockets and thin layers of clay. The speciied
requirement was for post-treatment CPT tests with minimum relative den-
sity of 75% using the Jamiolkowski et al. (1985) method and clay layers to
be for inferred undrained cohesion of at least 40 kPa.
As the groundwater was at 1.5 m depth, advance trials were performed to
conirm that the required performance could be achieved. The trials were
successful and the treatment proceeded using typically two, or locally three
depending upon visual monitoring, tamping passes with 8-tonne weight
dropped from heights of up to 8 m. Post-treatment CPT tests conirmed the
required sand density and that all weak clay pockets and layers had been
improved to minimum 60 kPa, typically 100 kPa. This success with the clay
layers is considered to be due their thicknesses being less than 200 mm, this
resulting in very short drainage paths for dissipation of the generated excess
pore-water pressures, combined with very careful attention to the recovery
periods between the tamping passes. The soils were considered unsuitable
for treatment by RIC.
The second site has received a number of visits to combine dynamic com-
paction to up to 6 m depth, with compaction grouting to chalk solution
features to up to 15 m depth, to ensure the integrity of the proposed basal
liner and leachate extraction system beneath substantial depth of landill.
The works were performed in the base of a former sand quarry where the
exposed solution features were clearly visible in their local extents. These
were then investigated by CPTs on nominal 3 m grid to assess whether
dynamic compaction alone was required or whether compaction grout-
ing had to be performed prior to the dynamic compaction. Some treat-
ment areas were as small as 10 m 2 , some as large as 1200 m 2 . The solution
features had been inilled by variable mixtures of sandy clays varying in
strength between very loose/soft and stiff, sandy silt, sand and medium-
dense gravelly sands. Areas of intact chalk were not subjected to treatment.
Post-treatment testing of the compaction grouting was performed by CPT
prior to the DC treatment. The dynamic compaction was then performed
82 Ground improvement

using three tamping passes with 8-tonne weight and drop heights of up to
12 m to achieve an undrained Young’s modulus of 50 MPa, as proven by
1.5 × 1.5 m large plate test loaded to 150 kPa.
The construction of a football stadium in Poland required treatment to
depths of 15 m and 10 m beneath the stadium and pitch areas, respectively.
A minimum CPT qc of 15 MPa was speciied below 1.0 m depth to permit
the adoption of foundation design pressures of up to 300 kPa.
The upper 3–4 m depth comprised peat and organic silt. These were to be
excavated and replaced by sandy soils. The soils to about 10 m depth then
comprised ine silty sands of about 10% silt content with pre-treatment qc
of about 5 MPa. The sands then became slightly coarser with up to 5%
silt content and qc varying between about 6 and 15 MPa. The sands also
contained thin silty layers and occasional traces of organics. Groundwater
was at about 1.5 m depth. As there were existing structures on several site
boundaries, the ground improvement design was based on vibro compac-
tion to the deeper sands followed by dynamic compaction to the upper
layers. Extensive vibro trials were irst performed, followed by dynamic
compaction trials. These revealed that wide vibro compaction grids alone
could achieve the speciied qc > 15 MPa at depth, but not to the shallower
layers. Closer and/or secondary vibro compaction grids to suitable depth
could then be performed or the dynamic compaction be designed to treat
to greater depth. Comparative costings revealed the optimum approach to
be relatively wide vibro compaction grids followed by dynamic compac-
tion using a 16-tonne weight dropped in free-fall from almost 20 m height
to improve the upper 8 m depth of the sands. The high groundwater table
would, however, have inhibited the eficient performance of such DC treat-
ment, with water entering the deep imprints before the full drop numbers
could be performed, thereby requiring phased treatment passes with suit-
able recovery periods between. Site levels were therefore raised by about 1.0
m to permit just two DC treatment passes for faster programme.
Post-treatment testing by CPT conirmed all speciied criteria had been
achieved (see Figure 3.9) with recorded settlements induced by dynamic
compaction, performed after the vibro compaction, of about 200 to 300 mm.

3.14 CONCLUDING REMARKS

The dynamic compaction method is a powerful tool when applied to suit-

able sites. A large database has been collated over the years to deine its
limitations and, more importantly, its capabilities, which can be utilised
with conidence. As with every specialist technique, the designer and con-
tractor performing this method of ground improvement must understand
these capabilities and limitations. Such understanding can only arise by
experience.
 Dynamic compaction 83

Comparison between pre-treatment, post-Vibro and post-DC result

qc (MPa)
0 10 15 20 30 40 50
0

6
Depth below ground level (m)

10

12

14

16
Before treatment 14 days after Vibro compaction (3.0m grid)
7 days after DC Acceptance criterion qc = 15MPa

Figure 3.9 Case history—DC after vibro compaction results.

84 Ground improvement

ACKNOWLEDGMENT

The author extends his appreciation to the various companies within the
Keller Group for permission to publish this information.

REFERENCES

BRE Digest 403 (1995). Damage to structures from ground-borne vibration. Building
Research Establishment, Watford.
BRE Report BR 458 (2003). Speciication for Dynamic Compaction. Building
Research Establishment, Watford.
BS 5228-1 and 2 (2009). Code of practice for noise and vibration control on con-
struction and open sites. British Standards Institution.
BS 7385 (1990). Part 1. Evaluation and measurement for vibration in buildings.
British Standards Institution.
Institution of Civil Engineers (1987). Speciication for Ground Treatment. Thomas
Telford, London.
D’Appolonia, E. (1953). Loose sands—their compaction by vibrolotation.
Symposium on Dynamic Testing of Soils, American Society of Testing
Materials, STP, p.156.
Dobson, T. and Slocombe, B.C. (1982). Deep densiication of granular ills. 2nd
Geotechnical Conference, Las Vegas, Nevada. April 1982.
Greenwood, D.A. and Kirsch, K. (1983). Specialist ground treatment by vibratory
and dynamic methods—state of the art report. Proceedings, Piling and Ground
Treatment for Foundations, London, pp. 17–45.
Jamiolkowski, M. et al. (1985). New developments in ield and laboratory testing
of soils. Theme lecture, 11th Int. Conf. on Soil Mechanics and Foundation
Engineering, San Francisco.
Kerisel, J. (1985). The history of geotechnical engineering up until 1700. Proceedings,
11th International Conference on Soil Mechanics and Foundation Engineering,
San Francisco (August 1985), pp. 3–93.
Leonards, G.A., Cutter, W.A. and Holtz, R.D. (1980). Dynamic compaction of gran-
ular soils. J. Geotech. Eng., ASCE 106 (GT1), pp. 35–44.
Lucas, R.G. (1995). Geotechnical Engineering Circular No. 1: Dynamic Compaction,
US Department of Transportation, Publication No. FHWA-SA-95-037.
Lundwall, N.B. (1968). The Saint George Temple. Temples of the Most High,
Bookcraft, Salt Lake City, Utah, Chapter 3, p. 78.
Mayne, P.W., Jones, J.S. and Dumas, J.C. (1984). Ground response to dynamic com-
paction. J. Geotech. Eng., ASCE 110 (6), pp. 757–774.
Menard, L. and Broise, Y. (1976). Theoretical and practical aspects of dynamic con-
solidation. Proceedings, Ground Treatment by Deep Compaction, Institution
of Civil Engineers, London, pp. 3–18.
Mitchell, J.K. and Katti, R.K. (1981). Soil improvement – state of the art report.
Proceedings, 10th International Conference on Soil Mechanics and Foundation
Engineering, Stockholm (June 1981), pp. 509–565.
 Dynamic compaction 85

Mitchell, J.K. and Solymar, Z.V. (1984). Time-dependent strength gain in freshly
deposited or densiied sand. J. Geotech. Eng., ASCE 110 (11), pp. 1559–1576.
Rollins, K.M. and Kim, J. (2010). Dynamic compaction of collapsible soils based on
US case histories. J. Geotech. Geoenviron. Eng., ASCE 133 (9), pp. 1178–1186.
Seed, H.B. and Idriss, I.M. (1971). Simpliied procedure for evaluating soil liquefac-
tion potential. J. Geotech. Eng., ASCE 97 (SM9), pp. 458–482.
Serridge, C.J. and Slocombe, B. (2012). Ground improvement. ICE manual of geo-
technical engineering. Eds. Burland, Chapman, Skinner and Brown. Chapter
84, pp. 1247–1269.
Serridge, C.J. and Synac, O. (2006). Application of the Rapid Impact Compaction
(RIC) technique for risk mitigation in problematic soils. Engineering Geology
for Tomorrow’s Cities. Eds. Culshaw, Reeves, Jefferson and Spink. Engineering
Geology Special Publication 22, Paper 294 (CD-ROM), Geological Society of
London.
Sharma, H.D. and Anirban, D. (2007). Municipal solid waste landill settlement:
Postclosure perspectives. Journal of Geotechnical and Geoenvironmental
Engineering, ASCE 133 (6), pp. 619–629.
Slocombe, B.C. (1989). Thornton Road, Listerhills, Bradford. Proceedings,
International Conference on Piling and Deep Foundations, London (May
1989), pp. 131–142.
West, J.M. (1976). The role of ground improvement in foundation engineering.
Proceedings, Ground Treatment by Deep Compaction, Institution of Civil
Engineers, London, pp. 71–78.
Chapter 4

Prefabricated vertical drains

Jian Chu and Venu Raju

CONTENTS

4.1 Introduction ................................................................................... 88

4.2 Consolidation theories and analysis................................................ 89
4.3 Design ............................................................................................ 99
4.3.1 Determination of design parameters .................................. 100
4.3.2 Properties of smeared soil .................................................. 104
4.3.3 Types of PVDs ................................................................... 106
4.3.4 Selection of PVDs .............................................................. 108
4.3.4.1 The discharge capacity ......................................... 109
4.3.4.2 Compatibility of the ilter with the soil to be
improved .............................................................. 111
4.3.4.3 The tensile strength of drain................................. 112
4.3.5 Pattern, spacing, and penetration depth for PVDs ............. 113
4.3.6 Settlement calculations ...................................................... 113
4.4 Construction ................................................................................ 115
4.4.1 Installation......................................................................... 115
4.4.2 Quality-control tests .......................................................... 120
4.4.2.1 Determination of the AOS .................................... 120
4.4.2.2 Permeability (or permittivity) of ilter ................... 123
4.4.2.3 Discharge capacity test ........................................ 123
4.4.2.4 Tensile strength tests............................................. 129
4.4.3 Measurement of penetration depth .................................... 131
4.4.3.1 Digitised PVD....................................................... 132
4.4.3.2 PVD with two wires ............................................. 133
4.4.3.3 PVD with one wire ............................................... 134
4.4.4 Field instrumentation and evaluation of performance........ 134
4.5 Design codes, standardisation, and speciications......................... 141
4.6 PVD for vacuum preloading ......................................................... 145
4.7 Case studies .................................................................................. 147
4.7.1 Reclamation and soil improvement for a slurry pond in
Singapore ........................................................................... 147

87
88 Ground improvement

4.7.2 Combined ill and vacuum preloading case........................ 153

4.8 Summary ...................................................................................... 158
Acknowledgments ................................................................................. 160
References ............................................................................................. 160

4.1 INTRODUCTION

Preloading is one of the most common ground improvement techniques for

soft clay. The modern application of the preloading method is combined
almost always with the use of vertical drains or prefabricated vertical
drains (PVDs). PVDs have been used successfully in many soil improvement
and land reclamation projects in the world (Hansbo, 1979, 2005; Holtz,
1987; Holtz et al., 1991; Balasubramaniam et al., 1995; Bergado et al.,
1990, 1993a, 1993b, 1996, 2002; Li and Rowe, 2001; Chu et al., 2004;
2009a; Choa et al., 2005; Bo et al., 2003; 2005; Arulrajah et al., 2004;
Indraratna et al., 2005; Seah, 2006; Kitazume, 2007; Varaksin and Yee,
2007). Therefore, the theories, design, and construction methods for PVDs
have become the core technical issues in the preloading or consolidation
method. In recent years, PVDs or their variations have also been used for
other purposes such as for dissipation of pore water pressures for liqueiable
sand (Towhata, 2008; Chu et al., 2009a) or in environmental engineering
for vapour extraction system (Schaefer et al., 1997; Collazos et al., 2002).
Depending on how a preload is applied, the preloading methods can be
subdivided into preloading using ill, preloading using vacuum pressure, and
combined ill and vacuum preloading methods, as described in Table 4.1. In
addition to preloading, PVDs have also been used for some other relatively
new methods such as dynamic consolidation for clays, which are also listed
in Table 4.1. In most of the applications, the main purpose of using PVDs
is to reduce the drainage path so that the time taken for the consolidation
of soft soil or the dissipation of excess pore water pressure can be substan-
tially reduced.
The practice of using vertical drains started with sand drains and then
evolved into PVDs. According to Hansbo (2004), the use of vertical sand
drains was irst proposed in 1925, and patented in 1926 by Daniel D.
Moran. A sand drain is formed in situ by placing sand directly into a bore-
hole or into a ‘‘sock’’ made of geosynthetic fabric or geotextile in a bore-
hole. One of the most well-documented case histories for the use of sand
drains is the test ield at Skå-Edeby in Sweden (Hansbo, 1960). However,
there are construction constraints in the installation of sand drains. PVDs,
also named band drains or wick drains, were introduced as a better alter-
native. The irst type of PVD was developed by Walter Kjellman in 1947
(Kjellman, 1948). It was made of wood and cardboard. Nowadays most
PVDs are made of corrugated plastic cores surrounded by geotextile ilters.
 Prefabricated vertical drains 89

The size of the PVDs has been standardised to a width of around 100 mm
and a thickness of 3 to 6 mm. Although PVDs are used for almost all the
projects, sand drains are still in use for some special projects. One example
is the Kansai International Airport where 400 mm diameter sand drains
were installed from an offshore barge into the soft seabed soils (Kitazume,
2007). PVDs made of natural products such as jute or coconut coir (Lee
et al., 2003) are also available.
A great deal of experience and development in both research and con-
struction have been accumulated in the past as a result of the extensive
use of PVDs. Two books devoted exclusively to the use of PVDs for soil
improvement have been published (Holtz et al., 1991; Bo et al., 2003).
A number of other books and reports with good coverage of PVDs have
also been published. These include Mitchell and Katti (1981), ICE (1982),
Jamiolkowski (1983), Akagi (1994), Bergado (1996), Moseley and Kirsch
(2004), Raison (2004), Hansbo (2005), Indraratna and Chu (2005), Chu
et al (2009a), and Chu et al. (2012).
This chapter intends to provide a practical guide to researchers and
practicing engineers who have to deal with design and construction issues
related to PVDs. It also tries to introduce briely the latest development and
technologies in the use of PVDs. Two case studies are also given to illus-
trate the real operation of soil improvement using PVDs.

4.2 CONSOLIDATION THEORIES AND ANALYSIS

The design of PVDs provides solutions to questions such as (1) what is

the drain spacing required to achieve a required degree of consolidation
within a given time; or (2) how long will it take to achieve a required degree
of consolidation for a given drain spacing and duration? To answer these
questions, consolidation theories are required.
The most fundamental consolidation theory is Terzaghi’s one-dimen-
sional consolidation equation, which can be written as

∂ue ∂ 2 ue kv
= cv and cv = (4.1)
∂t ∂z 2 mvγ w

where cv is the coeficient of consolidation of soil in the vertical direction,

ue is the excess pore water pressure, t is the real time, z is the position of
the soil element, kv is the coeficient of permeability of soil in the vertical
direction, mv = Δεv/Δσv’ is the coeficient of vertical compression, and γw is
the unit weight of water.
Equation 4.1 is derived under the assumptions that the dissipation of
water is only in the vertical direction (i.e., it can only be applied for con-
solidation with vertical low). However, with the use of PVDs, water lows
Table 4.1 Ground improvement methods where PVDs may be used

90
Method Description/Mechanisms Typical applications Advantages Limitations

Ground improvement
Preloading A. Preloading Preloading is a process to apply The method is Rate of consolidation can be The method may not be
methods using ill with surcharge load on to ground applicable to soils greatly accelerated. The applicable when the
vertical drains prior to the placement of having low construction time can be construction schedule is very
structure or external loads to permeability or when controlled by adjusting the tight or when the ground is
consolidate the soil until most the compressible soil spacing of the drain. so soft that vertical drains
of primary settlement has layer is thick. cannot be installed.
occurred.Vertical drains are
used to provide radial
drainage and accelerate the
rate of consolidation by
reducing the drainage paths.
B. Vacuum The method is the same as A, The method is 1). The method does not 1). This method causes
preloading except the surcharge is applicable to ground require ill material; inward lateral movement
with vertical applied using vacuum and consists of mainly 2). The construction and cracks on the ground
drains pressure. The vacuum saturated, low period can be shorter, surface, which may affect
pressure is usually distributed permeability soils. The as no stage loading is surrounding buildings or
through vertical drains. It also method can be used required; structures;
provides immediate stability when there is a 3). It may be more 2). The vacuum pressure is
to the system. The treated stability problem with economical than using limited to 50–90 kPa,
soil is enclosed by an air- and ill surcharge. This ill surcharge; depending on the system
watertight barrier in all method can also be 4) The vacuum brings adopted.
directions. used to extract immediate stability to
polluted ground pore the system.
water, if required.
C. Combined ill The method is a combination The same as for A 1). Construction time can 1). It is technically more
and vacuum of A and B when a surcharge and B. be much reduced as demanding than A and B;
preloading more than the limit of vacuum compared to staged 2). Data interpretation is
with vertical pressure is required. loading using ill also more complicated.
drains surcharge alone;
 2). The lateral movement
of soil can be
controlled by balancing
the amount of vacuum
and ill surcharge used;
3). The vacuum brings
immediate stability to
the system.
Dynamic D. Drainage This method improves the soil This method can be 1). This method makes the 1). The method may only
consolidation enhanced properties by combining the used to improve the application of DC work for cohesive soil
methods dynamic DC method with vertical bearing capacity of possible to ine-grained with relatively low
consolidation drains which facilitates the soft soil with low soil; plasticity index;
(DC) dissipation of pore water permeability. 2). The duration of soil 2). The compaction energy
pressure generated during improvement can be applied has to be within a
DC. reduced. certain limit, so that the
depth of improvement is
limited;
3). The technique has not
been fully developed.
Thus, the success of the
method cannot always be
guaranteed.

Prefabricated vertical drains

E. Dynamic This method improves the soil This method can be 1). The soil improvement 1). The method has not been
consolidation properties by conducting DC used to improve soft time can be reduced; fully established. Thus, the
(DC) and applying vacuum or clay or soft ground 2). Can be applicable to success of the method
combined with de-watering alternately for a with mixed soil. most types of soil. cannot always be
vacuum or number of times.The vacuum guaranteed;
de-watering facilitates a quick dissipation of 2). The depth of improvement
water pressure generated by is normally limited to 8 m.
DC.
Source: Modiied from Chu et al. 2009.

91
92 Ground improvement

mainly in the horizontal direction. In this case, the following radial con-
solidation theory is required (Barron, 1948):

 ∂2 u 1 ∂ue  ∂ue kh
ch  2e +  = and ch = (4.2)
 ∂r r ∂r  ∂t mvγ w

where ch is the coeficient of consolidation of soil in the horizontal direc-

tion, kh is the coeficient of permeability of soil in the horizontal direction.
Equation 4.2 was derived under the following two assumptions (Barron,
1948): (1) All vertical loads are initially carried by excess pore water pressure;
and (2) All compressive strains within the soil occur in the vertical directions.
When combining the vertical and horizontal low, the consolidation
equation becomes (Carillo, 1942):

∂ 2 ue  ∂ 2 ue 1 ∂ue  ∂ue .
cv + c h + = (4.3)
∂z 2
 ∂r
2 
r ∂r  ∂t
For ground improvement, the progress of consolidation of soil is evaluated
using the average degree of consolidation, U, of the compressible clay layer:
Sc (t)
U " 100% (4.4)
(Sc )ult
where Sc(t) is the consolidation settlement at a given time and (Sc)ult is the
ultimate consolidation settlement.
Using Terzaghi’s consolidation equation, Equation 4.1, and by assuming
the form of initial pore water pressure distribution, a relationship between
the average degree of consolidation and the time factor Tv = cvt/Hd2 can be
established as shown in Figure 4.1 or expressed approximately in equations
by curve itting. One of the closed-form equations is given by Sivaram and
Swamee (1977):

Sc (t) (4Tv /π )0.5

Uv = 100% = (4.5)
(Sc )ult [1 + (4Tv /π )2.8 ]0.179

where Uv is the average degree of consolidation due to vertical low.

For radial consolidation, the following solution was given by Barron
(1948) by assuming equal strain and all the other assumptions adopted for
Equation 4.1:

 − 8Th 
Uh = 1 − exp   (4.6a)
 F ( n  )
n2 (3n 2 − 1)
F(n) = 2 ln(n) − ≈ ln(n) − 0.75 (4.6b)
(
n −1 ) 4 n2
 Prefabricated vertical drains 93

0
0.1 m=∞
2 e –M2TV
UV = 1 – ∑
Average consolidation ratio UV

0.2 M2
m=0
0.3 (4TV /π)0.5
= 0.179
0.4 1 + (4TV /π)2.8

0.5
0.6 UV =
0.7

0.8
0.9
1
0.01 0.1 1 10
TV

Figure 4.1 Relationship between average degree of consolidation Uv and time factor Tv
derived based on solutions to Equation 4.1. (Redrawn from Bo, M.W., Chu, J.,
Low, B.K. and Choa, V. (2003). Soil Improvement: Prefabricated Vertical Drain
Technique, Thomson Learning, Singapore.)

ch t
Th = (4.7)
de2
de
n" (4.8)
dw

whereTh is the time factor, n is the ratio between the diameter of soil dis-
charging water into a vertical drain, de, and the diameter of the drainage
well or equivalent diameter of a vertical drain, dw, as shown in Figure 4 2.
In this igure, a and b are the thickness and width of band drain.
The solutions given in Equations 4.6 through 4.8 were developed for a
unit cell (i.e., a cylindrical column of soil surrounding a circular well, as
shown in Figure 4.2). A graphical illustration of the relationships between
Uh and Th for different n values is shown in Figure 4.3. The curve shown
in Figure 4.1 for vertical low is also plotted in Figure 4.3 for comparison.
However, it should be noted that a PVD band drain is not circular. It has
a typical width of 100 mm and thickness ranging from 3–6 mm. To use
Equations 4.6 through 4.8, a conversion to compute the equivalent drain
diameter, dw, is thus necessary. One conversion method based on an equal
perimeter has been proposed by Hansbo (1979) as:

2(a + b)
dw = (4.9)
π
94 Ground improvement

de
a

Equivalent cylindrical drain

dw

Band drain

Tributary clay cylinder

Figure 4.2 Barron’s equal-strain solution for radial drainage in a cylindrical cell. (Redrawn
from Bo, M.W., Chu, J., Low, B.K. and Choa, V. (2003). Soil Improvement:
Prefabricated Vertical Drain Technique, Thomson Learning, Singapore.)

Equal-strain consolidation with no smear

0

0.1 n = 100
40
0.2
20
0.3
15
0.4
Uv and Uh

n = 10
0.5
7
0.6
5
0.7
n = de / dw
0.8 Vertical flow
0.9 Horizontal flow

1
0.01 0.1 1

Figure 4.3 Relationships between Uh and Th for different n according to Equation 4.6
and Uv and Tv according to Equation 4.5. (Redrawn from Bo, M.W., Chu, J.,
Low, B.K. and Choa, V. (2003). Soil Improvement: Prefabricated Vertical Drain
Technique, Thomson Learning, Singapore.)
 Prefabricated vertical drains 95

The diameter of soil discharging water into a vertical drain, de, is depen-
dent on the drain spacing and the drain installation pattern. The de is calcu-
lated based on equivalent cross-section area. For PVDs installed in a square
grid pattern with a spacing s as shown in Figure 4.4a, de can be calculated as:

U de 2
s2 = , i.e., de = 1.128s (4.10)
4
For PVDs installed in a triangle grid as shown in Figure 4.4b, de can be
calculated as:

π d e 2 , i.e., d = 1.05s (4.11)

s 2 sin 60°= e
4
For relatively long PVDs, consolidation of clay is controlled by horizontal
drainage. However, for relatively short PVDs, both vertical and horizon-
tal drainage may contribute a fair proportion. In this case, the combined
degree of consolidation Uvh can be calculated using Carillo’s equation
(Carillo, 1942):

(1 − U ) = (1 − U ) × (1 − U )
vh v h
(4.12)

In this case, Uv and Uh should be calculated separately using Equations 4.5

and 4.6.
It should be pointed out that Equation 4.6 was derived by assuming that
the well resistance of the PVD can be ignored. If the well resistance has to
be considered, the F(n) equation in Equation 4.6b will have to be changed
into (Hansbo, 1981):

kh
F(n) ≈ ln(n) − 0.75 + π z(2l − z) (4.13)
qw

where l = longest drainage path along vertical drain; z = depth; and qw =

discharge capacity of PVD.

de

s s

de

(a) Square pattern. (b) Triangular pattern.

Figure 4.4 Patterns of PVD installations.

96 Ground improvement

It can be seen from Equation 4.13 that when qw is suficiently large, the
last term in Equation 4.13 will become very small so that the effect of well
resistance can be ignored. Some modern high-quality PVD products can
provide suficiently large qw. Therefore, well resistance may be ignored in
the design when PVDs with suficient large qw are used. Practically this
requires the qw value of PVDs to be speciically checked as part of the
quality-control process during the construction. It is thus important to
ensure that the PVDs are selected properly and the quality of the PVD
products is checked whenever PVDs are used for soil improvement (Chu
et al., 2004).

Working Example 4.1

PVDs were installed in a compressible clay layer of 10 m thickness
in a square pattern with a spacing of 2 m. The PVD used is 100 mm
wide and 4 mm thick. The coeficients of permeability of the clay in
the vertical and horizontal directions is 2.0 m 2 /year and 3.0 m 2 /year,
respectively. The boundary below the clay was impervious. Calculate
the degree of consolidation achieved in one year’s time.
Solution:
de = 1.128s = 1.128 × 2 = 2.256 m = 2256 mm
dw = 2(a + b)/π = 2(100 + 4)/3.14 = 66 mm
n = de /dw = 2256/66 = 34
F(n) = ln(n) − 0.75 = ln(34) − 0.75 = 2.78
Time factor due to radial drainage:
Th = cht/de2 = 3 × 1/2.2562 = 0.589
Degree of consolidation due to radial drainage:
Uh = 1 − exp[−8Th /F(n)] × 100% = 1 − exp[−8 × 0.589/2.78] = 82%
Time factor due to vertical drainage:
Tv = cvt/Hd2 = 2 × 1/102 = 0.02
Degree of consolidation due to vertical drainage:
Uv = (4Tv /π)0.5/[1 + (4Tv /π)2.8]0.179 100% = 16%
The combined degree of consolidation is:
From (1 − Uvh) = (1 − Uh)(1 − Uv), we can calculate Uvh = 85%

From Working Example 4.1, it can be seen that the consolidation due to
vertical drainage is normally small when the PVD is relatively long (say,
more than 10 m). Therefore, depending on the design situation, it is possi-
ble to design based on radial drain alone as a more conservative estimation.
Another commonly encountered design problem is to calculate the time
taken to achieve a certain degree of consolidation, Uvh, for a given PVD
installation scheme. When time t is not known, it is dificult to calculate
Tv and Th and thus Uv, Uh, and Uvh. There are three methods to solve this
problem.
 Prefabricated vertical drains 97

(1) Take Uh = Uvh and calculate t using Equation 4.6 by ignoring the
contribution of vertical consolidation.
(2) By a trial and error method. The t calculated in method 1 can be
used as the irst estimate to calculate Uvh. If Uvh is greater than the
assumed, then a smaller t can be used to calculate Uvh again until the
Uvh value matches the assumed.
(3) Use a special function ‘‘GoalSeek’’ in the Microsoft Excel to carry
out iterations automatically to obtain the time required to achieve
the required degree of consolidation. The detail of the third method
and Excel code that can be used for this purpose are provided in Bo
et al. (2003).

It should be pointed out that these solutions are provided for perfect
drain conditions (i.e., the installation of PVDs does not affect the soil prop-
erties). However, during the PVD installation process, the soil around the
PVDs is disturbed or smeared. The ‘smear’ effect comes from the compress-
ibility of soil and the disturbance to the soil structure during the inser-
tion and removal of the mandrel (see Section 4.3 for details). The zone in
which the soil is disturbed or smeared is called the smear zone, as shown in
Figure 4.5. The diameter of the smear zone, ds, varies from soil to soil and
is also affected by the size of the mandrel. Based on past studies (e.g., Holtz
and Holm, 1972; Hansbo, 1981, 1983; Indraratna and Redana, 1998;
Onoue et al., 1991; Hird and Moseley, 2000; Xiao, 2002), Bo et al. (2003)
have proposed that ds to be estimated as:

dw

Vertical drain

Smear zone
Perfect drain

de/2

Undisturbed
clay

ds

Figure 4.5 Analytical model of smear zone around vertical drain.

98 Ground improvement

ds = (4 to 7) dw or ds = (3 to 4) dm (4.14)

where dm = equivalent diameter of the mandrel, dw = equivalent drainage

diameter.
As a result of smear or sample disturbance, the coeficient of permeabil-
ity or coeficient of consolidation in the smeared zone is greatly reduced
compared to the intact soil. The reduction in permeability can be normally
taken as 2 or 3 (i.e., the permeability of the smeared soil, ks, is 2 to 3 times
smaller than that of the intact soil, kh):

kh = (2 to 3) ks (4.15)

However, the study of Bo et al. (2003) and Chu et al. (2004) indicates that
the reduction in permeability can be as large as 2 to 10 times depending on
the sensitivity of the soil.
By taking the smear effect into consideration, Barron (1948) and Hansbo
(1979, 1981) have derived another set of consolidation equations by assum-
ing an annulus of the smeared clay around the drain with a diameter of ds
and a permeability of ks:
 − 8Th 
Uh = 1 − exp  
)
 Fs ( n 
k 
Fs (n) ≈ ln(n) − 0.75 + ln(s)  h − 1 (4.16b)
 ks 

where s = ds /dw is the smear ratio.

It can be seen from Equation 4.16 that when s = 1 and k h = k s , so
when there is no smear effect, Equation 4.16 becomes identical to
Equation 4.6.

Working Example 4.2

PVDs were installed in a compressible clay layer of 10 m thick in a
square pattern with a spacing of 2 m. The PVD used is 100 mm wide
and 4 mm thick. The coeficient of permeability of the clay in the verti-
cal and horizontal directions is 2.0 m 2/year and 3.0 m 2/year, respec-
tively. The boundary below the clay was impervious. Assuming the
smeared ratio is 3 and the coeficient of permeability of the smeared
soil is 1.5 m 2/year. Calculate the degree of consolidation achieved in
one year’s time.
Solution:
In Working Example 4.1, we have already calculated de = 2.256 m =
2256 mm, dw = 66 mm, n = de /dw = 34, Th = 0.589, Tv = 0.02, and Uv =
16%. Using Equation 4.16b:
 Prefabricated vertical drains 99

Fs(n) = ln(n) – 0.75 + ln(s)(kh /ks − 1) = ln(34) − 0.75 + ln(3)(2 − 1) = 3.88

Degree of consolidation due to radial drainage:
Uh = 1 − exp[−8Th /Fs(n)] 100% = 1 − exp[−8 × 0.589/3.88] = 70%
The combined degree of consolidation can be calculated using Equation 4.12:
From (1 − Uvh) = (1 − 0.7)(1 − 0.16), we have Uvh = 75%.
Comparing the answers for the two working examples, it can be seen
that the degree of consolidation Uvh has reduced by 10% from 85% to
75% due to smear effect.

It should be pointed out that all the above analytical methods were
established based on Darcian low. For non-Darcian low, solutions have
also been provided by Hansbo (1997; 2001; 2004). However, the non-
Darcian low consolidation theory has not been widely used in practice.
Furthermore, the effect of non-Darcian low on one-dimensional consoli-
dation is negligible in the beginning of the consolidation process (Hansbo,
2004). As the non-Darcian low consolidation equations have been pre-
sented elsewhere (Hansbo, 1997; 2001; 2004; 2005), it will not be elabo-
rated in this chapter.
Note that the use of the analytical solutions presented above is restricted
to the assumptions of one-dimensional, linear-elastic, small strain behav-
iour of soil. The spatial variation of soil properties is also not taken into
consideration. For the modelling of two- or three-dimensional bound-
ary value problems with spatial variation of soil properties, such as con-
solidation under an embankment, the inite element analysis should be
used. Several numerical procedures have been developed for this purpose.
However, the coverage of these topics is beyond the scope of this chapter.
Interested readers are referred to Hird et al. (1992); Bergado et al. (1993b);
Chai et al. (1995; 2001); Indraratna and Redana (1997; 2000); Indraratna
et al. (2005); Rujikiatkanjorn et al. (2008); and Chu et al. (2012).

4.3 DESIGN

For the design of a soil improvement project using preloading and PVDs,
the following design and construction procedure can be adopted:

1. Conduct proper site investigation to establish the soil proile on site,

characterise the geotechnical properties of the soil, and determine
the design parameters.
2. Determine the depths that PVDs need to be installed and the pattern
of installation.
3. Select PVDs that meet the design speciications and design
requirements.
4. Calculate the drain spacing required to achieve the required design
speciication.
100 Ground improvement

5. Estimate the ground settlement and draw the surcharge placement

plan.
6. Install PVDs and carry out quality-control tests and inspections dur-
ing the PVD installation at predetermined intervals.
7. Design a ield instrumentation scheme, install instruments, collect
ield monitoring data, and monitor the soil improvement process.
8. Calculate the degree of consolidation and other design param-
eters used for design and check whether design speciications have
been met.

Some of these steps will be explained in detail in the following sec-

tions. The low chart illustrating the above procedure is also shown in
Figure 4.6.

4.3.1 Determination of design parameters

Once the consolidation theories and methods of analysis are in place, the
next step in the design process is to obtain soil parameters to feed into the
equations or computer software for analysis. This is not a simple task as
the determination of soil parameters is still one of the most challenging
tasks facing geotechnical engineers. We need to obtain a value for each soil
parameter, but few soil parameters are constant. For example, the coef-
icient of consolidation is assumed to be a constant in either Terzaghi’s or
Barron’s consolidation theory (i.e., Equation 4.1 or 4.2). However, in prac-
tice, the coeficient of consolidation for soft soil is not a constant. Its value
is affected by many factors, such as the overconsolidation ratio, the stress
state, the fabric of the soil, and even the method of determination (Holtz
and Kovacs, 1981; Chu et al., 2002). Therefore, the so-called engineering
judgment is sometimes required in deciding which value would be the most
appropriate. Good engineering judgment comes from good understanding
of soil behaviour and the past experience in dealing with similar types of soil
and geotechnical problems. The coeficient of permeability is another key
parameter required for vertical drain design. However, it happens that the
coeficient of permeability of soil is one of the most dificult soil parameters
to be determined. This is partially because the coeficient of permeability of
the soil has the widest range of variation among all the soil parameters. Its
value can vary from 10 –11 m/s for soft clay to 10 –3 m/s for sand and gravel,
a change of 108 times. Although the permeability of the soil that needs to
be treated with vertical drains is normally low, the error involved in the
permeability estimation can still range from 10 to 100 times. This is not
unusual as the permeability of the same soil can change 10 to 100 times
during the process of consolidation. An error of one order of magnitude in
permeability can result in an error of the same order of magnitude in the
time taken to achieve a speciic degree of consolidation based on Terzaghi’s
 Prefabricated vertical drains 101

( oil profile
Soil
( Geotechnical properties
investigation
( Design parameters

PVD depth
and pattern
(square or
triangular)

To check
( AOS of the filter
PVD type Quality
( Permeability of the filter
selection control test
( "! #'
( ensile strength

Drain spacing

Settlement
check and
surcharge
plan

Install PVD
Spacing and
and quality
depth check
control

Field Pore water pressure

instrumentation Settlement data

Check
degree of
consolidation

Check
specification
for design
adjustment

Figure 4.6 Design and construction procedures low chart.

102 Ground improvement

consolidation theory.* Therefore, it makes sense economically to conduct

some proper site investigation work and determine the soil parameters as
accurately as possible.
The speciic soil parameters that are required for the design of soil
improvement work involving vertical drains in soft clay include:

(1) The preconsolidation stress, σp′, and the overconsolidation ratio (OCR).
(2) The coeficient of consolidation in both horizontal and vertical direc-
tions, c h and cv.
(3) The coeficient of permeability in both horizontal and vertical direc-
tions, kh and kv.
(4) The diameter of the smeared zone, ds, and the permeability of the
smeared soil, ks.
(5) The coeficient of compressibility, C c, the coeficient of recompress-
ibility, C r, and sometimes the secondary compression index, C α , are
required for settlement estimation.
(6) The undrained shear strength, cu, and the undrained Young’s modu-
lus, E u, may also be required for analysing the stability of a dike or
the stability of a drain installation rig on soft clay.

As mentioned before, the values of ch and cv, or kh and kv change with

stress state or OCR. One has to accept the fact that neither the coeficient
of consolidation nor the coeficient of permeability of soil is a constant. As
such, the selection of those parameters has to be based on its in-situ stress
conditions and the anticipated stress changes. Therefore, it is also necessary
to establish relationships between the coeficient of permeability and void
ratio, and relationships between the coeficient of consolidation, and the
stress state. A proper site investigation should be planned not only to deter-
mine the soil parameters but also to understand how the soil parameters
vary with stress and loading conditions.
Generally the consolidation parameters of soil can be determined using
laboratory tests, in-situ tests, back-calculation from ield measurements, or
a combination of them. In laboratory tests, the stress states and drainage
conditions can be deined precisely and the variation of soil parameter with
stress and consolidation process can be evaluated. However, the results
are usually affected by sample disturbance. It is also time consuming to

*
TH d2 mvγ w ,…
t = = TH d2
cv kv
– dkv ,
dt = TH mv γ w2
d
kv2
dt TH mvγ w – dkv 1 – dkv – dkv
2
d
= = cv =
t t kv kv cv kv kv
 Prefabricated vertical drains 103

conduct laboratory consolidation tests. In-situ tests are normally relatively

quick to conduct and therefore are more useful than laboratory tests in
identifying the soil proile and characterising the soil behaviour over a large
extent. However, in in-situ tests, the stress and drainage conditions are
generally not well deined. The data interpretation from physical measure-
ments to soil parameters are sometimes based on arbitrary assumptions or
correlations which are established for a speciic type of soil only. Therefore,
when in-situ tests are adopted, laboratory tests may still be required to
verify the assumptions and check the correlation relationships. The back-
calculation from ield measurements can provide a good check on the selec-
tion of design parameters. However, the back-calculated value is only a
factored parameter. It relects not only the soil property, but also other fac-
tors, such as the disturbance to the soil during construction.
The types of laboratory and in-situ tests that are suitable to the deter-
mination of consolidation properties are summarised in Table 4.2. The

Table 4.2 Types of tests for measurement of consolidation properties

Parameter
Type of test Name of test determined Remarks
Laboratory Oedometer test cv, kv (indirect Need high-quality
tests measurementi), Cc, Cr, ‘undisturbed’
σp′, and Cαii samples.
Rowe cell test ch and kh (directiii or
indirect
measurement)
Other consolidometers ch and kh (direct or
indirect
measurement)
In-situ tests Piezocone dissipation ch and kh (indirect Based on pore water
test (CPTU) measurement) pressure dissipation.
Pressuremeter or ch and kh (indirect Based on lateral
self-boring measurement) pressure change or
pressuremeter (SBPM) pore water pressure
test dissipation.
Flat dilatometer test ch and kh (indirect Based on lateral stress
(DMT) measurement) change.
Field permeability test kh (direct Using a piezometer.
(e.g., BAT permemeter) measurement)
Back- Based on pore water ch (factored value) Using piezometers.
analysis pressure
measurements
Based on settlement ch (factored value) Using settlement
measurements. gauges.
i In this case, kv is calculated based on the value of cv.
ii When secondary consolidation is measured.
iii kh is measured directly as part of the consolidation test.
104 Ground improvement

settlement prediction for projects using vertical drains is the same as those
without the use of vertical drains. Those methods are covered in many
textbooks (e.g., Holtz and Kovacs, 1981). As far as land reclamation or the
other similar types of geotechnical problems (where the extent of load is
much greater than the thickness of the compressible layer) are concerned,
the settlement predicted using the one-dimensional analysis and param-
eters determined by laboratory tests is reasonable although not always reli-
able. Ground settlement should always be monitored as part of the soil
improvement works.
Different types of laboratory and in-situ tests that are suitable for the
determination of consolidation properties are discussed in detail in Chu
et al. (2002) and Bo et al. (2003). One example for the determination of the
c h values for the intact Singapore marine clay at Changi by different methods
is presented in Figure 4.7. Among the tests shown in Figure 4.7, the in-situ
tests and the Rowe cell test measure the c h values, whereas the conventional
oedometer test measures cv. The c h back-calculated based on the settlements
measured at different elevations are also presented in Figure 4.6. In the
back-calculation, the ultimate settlement was estimated based on Asaoka’s
method (Asaoka, 1978). As shown in Figure 4.6, the back-calculated c h
values were lower than the c h values determined by either laboratory or
in-situ tests. Similar observations have been made at other sites in the
Singapore marine clay (Chu et al., 2002) and by Balasubramaniam et al.
(1995) for the Bangkok clay. Prefabricated vertical drains were installed
at those sites at a close spacing. It implies that when vertical drains are
used in soft marine clay, the overall c h value of the soil will be lower than
the c h value of the intact soil. This could be due to the smear effect to soil
induced by the installation of vertical drains. The effect of disturbance can
be relatively large particularly when the drains are installed at a close spac-
ing (Chu et al., 2002).
It has been generally observed from the comparisons made in Figure 4.7
and the other cases that:

(a) The c h of soft clay determined by the Rowe cell test is generally 2–4
times larger than the cv by the conventional oedometer test, relecting
the anisotropic nature of the soil.
(b) The CPTU dissipation test tends to agree reasonably well with that
from the Rowe cell tests. Therefore, CPTU can be a good tool for c h
determination if it is calibrated properly.

4.3.2 Properties of smeared soil

With the substantial reduction in recent years in the costs involved in verti-
cal drain products and installation, there is a tendency to use closer drain
spacing. However, when the drains are installed too close to each other,
 Prefabricated vertical drains 105

0
SBPM

DMT

10
CPTU

Ch(Lab)
Depth (m)

20 Cv(Lab)

Field

30

40
0 5 10 15 20
Cv or Ch (m2/yr)

Figure 4.7 Comparison of ch proile of Singapore marine clay measured by different

methods.

the beneit resulting from use of more PVDs may be overshadowed by

the increased smear effect due to PVD installation. Although Equation
4.14 has been proposed to estimate the diameter of the smear zone, ds,
it is dificult to quantify ds accurately as the value ds is affected by many
factors, such as the shape, the size of the mandrel, the type of soil, and
the sensitivity of soil. Several studies have been conducted to determine ds
and the effect of smear zone on the consolidation of soil. Hansbo (1981,
1997) estimated d s = (1.5 ~ 3.0)dw. This relation has been commonly used
in design. Based on a laboratory study and back analyses, Bergado et al.
(1991) proposed that ds = 2dw could be assumed. Indraratna and Redana
(1998) observed from some model tests that the smear zone could be as
large as ds = (4 ~ 5)dw. Holtz and Holm (1972) also suggested that ds be
equal to two times the equivalent diameter of the mandrel. The studies of
Xiao (2002) and Bo et al. (2003) indicated that the smear zone could be as
large as 4 times of the size of the mandrel or 5–7 times the equivalent diam-
eter of drain. The dimension of a small rectangular mandrel is 120 mm by
60 mm. This is equivalent to a diameter of 115 mm using Equation 4.9. The
equivalent diameter of a drain is 66 mm. Therefore, the size of the smear
zone can be as large as 460 mm. If the drain spacing is 1 m, it means almost
everywhere the soil is disturbed. As smear can cause a signiicant reduction
106 Ground improvement

in the permeability and the coeficient of consolidation of soil, it can be

counterproductive when the drain spacing is too close. This is particularly
the case when the soil is sensitive.
The parameter, ks, is normally estimated using a reduction ratio, ks/kh.
A value in the range of 2 to 6 has been suggested by various researchers
(Hansbo, 1981; Onoue et al., 1991; Indraratna and Redana, 1998; Hird
and Moseley, 2000). Hansbo (1997) also proposed the ratio be put equal
to the ratio of the permeability in the horizontal direction to that in the
vertical direction.

4.3.3 Types of PVDs

PVD products have become quite standard in terms of size and perfor-
mance. The typical cross-section is 100 mm wide and 3–6 mm thick. A
PVD normally consists of a core and ilter made of different types of mate-
rials. Classiied in terms of design, the PVDs available in the market can be
classiied into three different types. The irst is the ordinary type of band
drains where the core and ilter are itted loosely and are separable. Some
typical forms of PVDs are shown in Table 4.3. The photos of three PVD
samples are shown in Figure 4.8. The second type is the so-called integrated

Table 4.3 Ordinary types of PVD

Core Filter type Method of assembly

Separated Filter
ilter and jointed to
Cross-section Description sleeve core

Corrugated groove Nonwoven Mebradrain

fabric MD 7007

Ribbed groove Synthetic Mebradrain

iber MD88
Monoilament Needle- Colbond
punched CX1000
nonwoven
fabric
Double Cuspated Nonwoven Flodrain
fabric
Studded (one side) Nonwoven Alidrain ST
fabric
Studded (two sides) Nonwoven Alidrain DC
fabric
 Prefabricated vertical drains 107

Figure 4.8 Samples of ordinary types of PVDs.

PVD where the core is adhered to the ilter, as shown in Figure 4.9. This
type of PVD offers a number of advantages over the ordinary type of PVDs.
For examples, it offers a higher discharge capacity and tensile strength as
discussed in detail by Liu and Chu (2009). The third type is circular PVDs.
A picture of it is shown in Figure 4.10. The circular PVD has better resis-
tance to buckling and has been used exclusively for vacuum preloading
projects (Chu et al., 2009).
It should be mentioned that there are different types of drains, such
as electric vertical drain with a metal foil embedded in the drains as
anodes and cathodes for electro-osmosis (Shang, 1998; Bergado et al.,
2000; Karunaratne, 2011). There are also PVDs for geoenvironmental
use. For example, PVDs have been used for a vapour extraction sys-
tem (Schaefer et al., 1997; Collazos et al., 2002). There have also been

100 ± 2 mm
5 ± 1 mm

Figure 4.9 Integrated type of PVDs. (Redrawn from Liu, H.L. and Chu, J. (2009).
Geotextiles and Geomembranes, Vol. 27, No. 2, pp. 152–155.)
108 Ground improvement

Figure 4.10 Circular type of drain used for some vacuum preloading projects.

studies to produce biodegradable PVDs using biodegradable plastics

(Park et al., 2010).

4.3.4 Selection of PVDs

The uncertainties involved in the design of PVDs include

1. The bias in the analytical model

2. Uncertainties or errors in the parameters entered to Equations 4.6
and 4.16
3. Variations in the quality of drain materials and uncertainties involved
in installation

As far as the analytical model is concerned, Equation 4.6 or 4.16

represents only an oversimpliication of the real three-dimensional
consolidation process. However, a comparison of the solutions with a
sophisticated elasto-plastic inite element analysis (Xiao, 2002) shows
that the former can normally provide a good estimation of the degree
of consolidation, Uh, although it may not be able to predict the pore
water pressure distribution reliably. Therefore, for normal design pur-
pose, the solutions given by Equation 4.6 or 4.16 would be adequate for
most cases.
As mentioned before, in Equations 4.6 and 4.16 the effect of well resis-
tance is not considered. This is a reasonable assumption if the PVDs used
can provide suficient discharge capacity, qw. Although good PVDs can
meet this requirement easily, it is important in practice to have a qual-
ity-control procedure in place so that the quality of PVDs used can be
checked to ensure qw of the PVDs is adequate and all the other require-
ments are met.
 Prefabricated vertical drains 109

The quality and suitability of the drains play a key role in the whole
soil improvement scheme involving PVDs. Different design situations
require different types of PVDs. For example, it is not necessary to use a
vertical drain with a high discharge capacity value if the drain is short.
The drain ilter should also match the soil type. The unit price of verti-
cal drain is another important consideration besides meeting the design
requirements. A considerable saving can be achieved without sacriicing
the performance of the drain, if the control factors for a vertical drain can
be identiied and the design requirements are speciied accordingly. The
factors that control the selection of vertical drain, apart from the cost,
are as follows.

4.3.4.1 The discharge capacity

It should be pointed out that the well resistance is controlled not only by
the discharge capacity of the drain, qw, but also by the permeability of soil,
ks, and the longest discharge length, lm , as can be seen from the last term
of Equation 4.13. To evaluate the eficiency of drain in discharging water,
a dimensionless parameter, the so-called discharge factor is deined (Mesri
and Lo, 1991):

qw
D" (4.17)
khl m2

Equation 4.17 relects the fact that the larger the k h or the longer
the drain, the larger the discharge capacity is required. The effect of
well resistance can be evaluated using this discharge factor. Based on a
numerical study, Xie (1987) established that in order to meet the assump-
tion that the well resistance can be ignored, the following condition must
be met:

π kh 2
l < 0.1 (4.18)
4 qw m

This requires the discharge factor, D, to be:

qw (4.19)
D= ≥ 7.85
khl m2

Therefore, the required discharge capacity after applying a factor of safety

to consider all the inluencing factors on discharge capacity including buck-
ling becomes:

qreq v 7.85Fs khl m2 (4.20)

110 Ground improvement

where qreq = the required discharge capacity, Fs = factor of safety, normally

Fs = 4 ~ 6.
The condition given in Inequality (4.20) is consistent with the thresh-
old discharge factor of 5 speciied by Mesri and Lo (1991). Inequality
(4.20) deines the dependence of qw on k s and lm . The relationship among
qw, k s , and lm for Fs = 5 for negligible well resistance has been plotted in
Figure 4.11.
Based on Inequality (4.20), lm has the most signiicant inluence on the
required discharge capacity. If we take Fs = 5, ks = 10 –10 m/s, and lm = 25 m,
then qw = 2.45 × 10 –6 m3/s, or 82 m3/yr. If lm = 50 m instead of 25 m, then
qw = 9.81 × 10 –6 m3/s, or 327 m3/yr. Nowadays, most of the drains can pro-
vide such a qw value even under a buckled condition. On the other hand,
permeability can also have a great effect when it is not determined accu-
rately. Take the previous case for example, if ks = 10 –9 m/s instead of 10 –10
m/s, then qw = 98.1 × 10 –6 m3/s, or 3,270 m3/yr. In this case, some drains
will not be able to meet the requirement.
For the Changi East land reclamation project, the drains used were up to
50 m long. The permeability of the soil normally ranges from 10 –11 to 10 –10
m/s. Using Inequality (4.20), qw required should not be larger than 9.81 ×
10 –6 m3/s. qw = 10 × 10 –6 m3/s was adopted as the speciication for the buck-
led drain at a pressure of 350 kPa for the Changi East land reclamation
project. Back-calculations of qw were made using the ield monitoring data
(Bo, 2003). The back-calculated qw was normally much smaller than the
speciied value, but some values were as high as 5 × 10 –6 m3/s (Bo, 2004).
As the ield measurements agree reasonably well with qreq calculated from

1000

k = 10–9 m/s
100
Required discharge capacity

10–10 m/s
(× 10–6 m3/s)

10

10–11 m/s
1

0.1

0.01
0 20 40 60 80
Maximum drain length (m)

Figure 4.11 Required discharge capacity as a function of PVD length and permeability of soil.
 Prefabricated vertical drains 111

Inequality (4.20), the condition speciied in Inequality (4.20) appears to be

adequate. It also indicates that the use of a factor of safety of 5 in Inequality
(4.20) is reasonable, although a slightly lower value can also be adopted.
This is why the factor of safety to be adopted in Inequality (4.20) is sug-
gested to be in the range of 4 to 6.
As the discharge capacity required is controlled by the permeability of
the soil and the length of vertical drain, in theory, the discharge capacity
required varies from project to project, and even varies from point to point
even within the same project. Therefore, the discharge capacity speciied
for one project may not be applicable to another even when the soil condi-
tions in the two projects are similar.

4.3.4.2 Compatibility of the filter with

the soil to be improved
The pore size or the apparent opening size (AOS) of the ilter should meet
the ilter design criteria. On one hand, the AOS has to be small enough to
prevent the ine particles of the soil from entering the ilter and the drain.
On the other hand, the AOS cannot be too small as the ilter has to provide
suficient permeability. The two key parameters that indicate the quality of
the ilter are the AOS and the cross-plane permeability of the ilter. Some
criteria for AOS have been proposed by several researchers. A commonly
used criterion is given by Carroll (1983):

O95 ≤ (2 ~ 3) D85 (4.21a)

and

O50 ≤ (10 to 12) D50 (4.21b)

where O95 is the AOS of ilter, O50 is the size which is larger than 50%
of the fabric pores, and D85 and D50 refer to the sizes for 85% and 50%
of passing of soil particle by weight. O95 ≤ 0.075 mm, or 75 μm, is often
speciied for PVDs.
The D85 of the Singapore marine clay was in the range of 0.01–0.02
mm (Bo et al., 2003). The O95 of the PVD ilter speciied for the Changi
land reclamation project was less than 0.075 mm. An AOS of 0.075 mm
exceeded the AOS speciied by Inequality (4.21a). This was permitted for
the following two reasons. First, PVDs with such O95 had been used suc-
cessfully in the previous land reclamation projects in Changi (Choa et al.,
1979). Second, there were not many types of PVDs which offer an O95
much smaller than 0.075 mm available in the southeast Asian market. To
verify whether the ilter was adequate, some drains had been pulled out
from the site at the end of the project. The inner side of the drain was quite
112 Ground improvement

clean. The same was observed from the drains used for some long-term
consolidation tests in the Singapore marine clay (Chu and Choa, 1995).
Therefore, the ilter criterion set by Inequality (4.21a) appears to be too
conservative for the Singapore marine clay based on this study. A more
relaxed criterion: O95 ≤ (4 ~ 7.5) D85, may be applicable to the Changi East
land reclamation project. For the Bangkok clay, a more relaxed criterion for
O95 has also been proposed (e.g., Bergado et al., 1993a).
The permeability of the ilter is normally required to be at least one
order of magnitude higher than that of the soil. Considering the clog-
ging effect, a much higher permeability should be required for the ilter.
Nevertheless, even more stringent requirements on the permeability of il-
ter can be met easily as most of the PVDs have a ilter permeability higher
than 10 –4 m/s, which is far greater than what is required. For example,
the permeability of the Singapore marine clay was in the order of 10 –10
m/s. The permeability of the ilter of the PVDs used for the Changi land
reclamation project was higher than 10 –4 m/s. This is 105 times greater.
The thickness of the ilter is another consideration. Normally the thicker
the ilter, the better it becomes, given other conditions the same. Based on
Wang and Chen (1996), the mass to area ratio should be generally larger
than 90 g/m 2 .

4.3.4.3 The tensile strength of drain

PVDs should have adequate tensile strength so that it can sustain the tensile
load applied to it during installation. Therefore, the strength of the core,
the strength of the ilter, the strength of the entire drain, and the strength
of the joint need to be speciied, normally at both wet and dry conditions.
According to Kremer et al. (1983), a drain must be able to withstand at least
0.5 kN of tensile force along the longitudinal direction without exceed-
ing 10% in elongation. It is quite common nowadays to specify the tensile
strength of the whole drain at both dry and wet conditions to be larger than
1 kN at a tensile strain of 10%. The same criterion was used for the Changi
East land reclamation project. The spliced drain is also required to have
strength comparable to that of the original drain. This criterion appeared
to be satisfactory for the PVDs used for the project. However, one factor
that is often neglected is that some drains can have permanent necking once
it is stretched. Such a necking reduces the discharge capacity. Therefore, the
amount of elongation and necking should also be observed and reported
during a tensile strength test. For the PVDs used for the Changi East land
reclamation project, necking only became obvious in the tensile strength
test after the tensile strain exceeded 20%, which is unlikely in the ield
condition.
It should be pointed out that although various methods and equipment
have been used for the determination of the properties of PVD, it may
 Prefabricated vertical drains 113

not be possible to compare directly the values provided by suppliers. Even

within the same method, the values measured can still vary depending on
the testing procedure. Furthermore, the discharge capacity is dependent
on hydraulic gradient. Thus, one can only compare the discharge capac-
ity when the values are measured using the same method and at the same
hydraulic gradient. Unfortunately, this is seldom the case. Under such cir-
cumstances, it is necessary to conduct one’s own tests to determine the
discharge capacity in the same way for all the drains concerned.

4.3.5 Pattern, spacing, and penetration

depth for PVDs
PVDs are installed in either square or triangular patterns. In theory, the tri-
angular pattern is slightly more cost effective. However, a square pattern is
simpler for layout and installation. The spacing for PVDs can be calculated
using the theories presented in Section 4.2. In theory, the closer the drain
spacing, the faster the rate of consolidation. However, when the spacing of
PVDs is too small, the smear effect becomes more signiicant as discussed
in the earlier section.
Normally PVDs should be installed through the entire depth of the
compressible soil layer. However, if the load is applied over a limited area,
such as a narrow embankment, the majority of the load may be distributed
within a certain depth only. In this case, it is not necessary to install PVDs
through the entire depth of the compressible soil layer. One example of
such a case is given by Yan et al. (2009).
Prefabricated vertical drains shall be located, numbered, and staked by
the surveyor using a baseline and benchmark indicated by the engineer.
The as-built location of the PVDs shall not vary by more than 150 mm
from the planned locations. During installation the depth and the length of
the drain installed at each location should be determined. The drain mate-
rial shall be cut neatly at its upper end with 100 to 200 mm protruding
above the working surface.

4.3.6 Settlement calculations

During the preliminary design stage, the consolidation settlement is nor-
mally calculated based on a one-dimensional settlement analysis using
data obtained from one-dimensional consolidation tests. One method of
using Microsoft Excel spreadsheet for settlement calculation by considering
staged loading is shown in Bo et al. (2003). As the one-dimensional settle-
ment calculation has been covered in many papers and textbooks, it will
not be elaborated on in this chapter. However, comments related to settle-
ment calculation follow.
114 Ground improvement

First, the settlement prediction using oedometer testing data can only be
as reliable as the data. For this reason, it is important to obtain good one-
dimensional consolidation test results. This requires good quality samples
to be used. It is well known that the sample quality affects the determina-
tion of preconsolidation stress and compression index (Holtz and Kovacs,
1981). A small error or uncertainty in the determination of preconsolida-
tion stress can cause a large variation in the settlement. Furthermore, it
should be noted that the preconsolidation stress changes with the method
of testing and rate of loading when the constant-rate-of-strain method is
adopted. Second, the preconsolidation stress and vertical stresses in soil
also vary with depth. It is therefore necessary to determine the precon-
solidation stresses at different depths and calculate the consolidation
settlement using the subdivision method unless the compressible layer is
relatively thin. Third, the surcharge applied by ill can change with the
settlement of the ground as part of the ill may submerge into water and the
effective surcharge load can thus be reduced. One such example is given by
Bo et al. (1999). Finally, it should be pointed out that this one-dimensional
approach is only reasonable when the extent of the load applied is much
larger than the thickness of the compressible soil layer such as in a land
reclamation project. Even in this case, settlement prediction has never been
an easy task, as elaborated by Duncan (1993). For this reason, settlement
prediction using ield monitoring data is essential for preloading projects
using PVDs.
Sometimes, the settlement due to secondary compression needs to be
estimated too. The conventional method of estimating the secondary com-
pression is to use

Δe = C α (Δlog t) (4.22)

where Δe is the additional compressibility due to secondary compression,

C α is the secondary compression index, and t is time. Mesri and Castro
(1987) observed the following relationship for natural medium plasticity
index clays:

C α /C c = 0.04 ~ 0.05 (4.23)

where C c is the compression index of soil.

Equations 4.22 and 4.23 can be used as a preliminary assessment for
the order of magnitude of secondary compression. A few advanced meth-
ods have also been proposed for the estimation of the secondary compres-
sion (e.g., Yin and Graham, 1999). However, the prediction of secondary
compression of soil is even more dificult than that for primary consoli-
dation. Part of the reason could be because the secondary compression
may not even be separated from the primary consolidation (Leroueil,
 Prefabricated vertical drains 115

1988). If secondary compression has to be considered in the design, the

intended preloading should be designed with an additional objective to
reduce the secondary settlement. Studies have shown that (Hight, 1999;
Mesri et al., 2001) the secondary compression index is much smaller
when soil is overconsolidated. Therefore, one effective way to reduce the
effect of secondary compression is to preload the soft ground to a con-
solidation stress higher than the anticipated working load. Then the soil
can be in an overconsolidated state during the working stage and thus
the settlement due to second compression will be insigniicant under the
working load.

4.4 CONSTRUCTION

4.4.1 Installation
The PVD installation procedure is illustrated in Figure 4.12. A PVD instal-
lation rig is normally used to penetrate a metal mandrel with PVD inside.
The PVD installation rigs used on land can be classiied into three types:
(1) static push-in type for normal ground (see Figure 4.13 for an example);
(2) vibratory drive-in type for irm to stiff ground or soft soil with a hard
crust; (3) light rigs on trucks for PVD installation on very soft ground, as
shown in Figure 4.14. For offshore or above-water PVD installation, a raft
or a PVD installation barge has been used. A raft such as the one in Figure
4.15 is only workable in quiet, shallow water. The barge is required in the
relatively deep water of an offshore environment. The PVD installation

Figure 4.12 PVD installation procedures. (Redrawn from http://www.americanwick.

com/uploads/documents/WICKDrainBrochure1.pdf.)
116 Ground improvement

Figure 4.13 A static push-in type of PVD installation rig.

barge shown in Figure 4.16 can install 12 PVDs at the same time. It has
been used for a port project in China (Yan et al., 2009).
The mandrel used for PVD installation is normally much larger than a
PVD in terms of cross-section areas as it has to be strong enough to pre-
vent it from bending or buckling. There are four different types of man-
drels according to the shape of the cross-sections: rhombic, rectangular,

Figure 4.14 A light-weight truck-supported installation rig. (Courtesy of S.W. Yan.)

 Prefabricated vertical drains 117

Figure 4.15 A loating raft for PVD installation in shallow water (From Chu, J., Bo,
M.W. and Arulrajah, A. (2009). Soil improvement works for an offshore
land reclamation. Proceedings of the Institution of Civil Engineers—Geotechnical
Engineering, Vol. 162, No. 1, pp. 21–32.)

square, and circular. The irst two are more commonly used. A picture of
their cross-sections is shown in Figure 4.17. The typical dimensions are
120–145 mm long, 60–mm wide, and 10–mm thick for rectangular cross-
section mandrel and 120–145 mm long, 50–85 mm wide, and 5–15 mm
thick for rhombic ones. It should be noted that the smear effect is greatly
affected by the cross-section area of the mandrel. Therefore, the cross-
section of the mandrel should be as small as possible in order to reduce the

Figure 4.16 A PVD installation barge. (From Yan, S.W., Chu, J., Fan, Q.J., and Yan, Y.
(2009). Building a breakwater with prefabricated caissons on soft clay.
Proceedings of ICE, Geotechnical Engineering, Vol. 162, No. 1, pp. 3–12.)
118 Ground improvement

Figure 4.17 Cross-section of mandrels. (From Bo, M.W., Chu, J., Low, B.K. and Choa,
V. (2003). Soil Improvement: Prefabricated Vertical Drain Technique, Thomson
Learning, Singapore.)

Figure 4.18 A bent mandrel coming out of the ground.

 Prefabricated vertical drains 119

Anchor bar

150 mm
12 mm

200 mm

Staple

Anchor plate
150 mm

80 mm
190 mm

Figure 4.19 Typical designs of anchors. (Redrawn from Bo, M.W., Chu, J., Low, B.K.
and Choa, V. (2003). Soil Improvement: Prefabricated Vertical Drain Technique,
Thomson Learning, Singapore.)

smear effect. Nevertheless, a mandrel must have enough stiffness to ensure

the verticality of the drain. A mandrel does bend when it hits a piece of
rock or a hard stratum. A mandrel bent inside the soil and coming out
behind the installation point is shown in Figure 4.18.
The PVD is pulled down together with the mandrel using an anchor bar
or anchor plate as shown in Figure 4.19. The PVD is ixed to an anchor bar
or plate using folding and staples. Two pictures illustrating the operation
are shown in Figure 4.20.

Figure 4.20 Use of an anchor plate for PVD installation. (Courtesy of J. Han.)
120 Ground improvement

PVDs come in rolls and sometimes have to be connected. Some details of

splicing for two different types of PVDs as used for a reclamation project
in Singapore are shown in Figure 4.21a and b. A minimum overlapping of
300 mm is normally required.

4.4.2 Quality-control tests

As several million metres of PVDs can be used in even a normal-sized soil
improvement project and the drains are produced over a period ranging from
a few months to a few years (probably at different factories), it is essential to
conduct quality-control tests on site to check the consistency of the products.
A list of speciications for PVDs used for different projects in different coun-
tries is given in Table 4.4. It is clear that quite different parameters and con-
trol values are used in the speciications for different countries. Although it is
better to check as many properties as possible, it may not be cost effective or
necessary to check on all the properties that appear in Table 4.4. Based on the
discussion in the preceding section, we identify the parameters that should be
checked for quality control for PVDs: the quality of the ilter, the discharge
capacity, apparent opening size (AOS), the tensile strength of the drain, as
well as the physical properties of the PVDs such as dimensions, weight to
area ratio, etc. The next question is how to conduct the quality-control tests.
Various methods have been proposed for measuring the various properties of
PVDs in the past. However, different tests or even the same test using differ-
ent control conditions will give different values. It should be pointed out that
the speciications given different PVD suppliers were measured using differ-
ent testing methods. Therefore, one cannot compare the values provided by
the PVD suppliers directly. For this reason, the end users should adopt a set
of testing methods for conducting not only quality-control tests, but also
comparing and selecting PVD products. The set of testing methods should
be simple enough that they can be carried out easily in any soil laboratory.
One such set of tests has been developed by Chu et al. (2004) and will be
described briely as follows. The basis for the selection of the suitable control
values for speciications is also discussed in the following.

4.4.2.1 Determination of the AOS

One of the common methods used to measure the AOS is to conduct sieve
analysis using standard beads. This method is applicable to AOS larger
than 40 μm. Standard ASTM D4751 (2004) is normally followed in con-
ducting this test, except that the diameters of the silicon beads used in the
test range from 40–170 μm, instead of 75–170 μm. The tests are conducted
under a relative humidity of 60% and a temperature of 20°C. The percent-
age of passing of the silicon beads is measured. The AOS of the ilter is
determined as a correlation with the percentage.
 (a) (b)

Open the filter backward 300mm of old drain as shown below:
Insert the new drain into the old drain:

300mm 300mm
New roll Old roll New roll Old roll

Inner core
Filter

Insert the opened core into the new drain and ensure smooth overlap:
Overlapped length should be 300mm

New roll 300mm New roll 300mm Old roll

Old roll

Overlapped inner core Overlapped inner core

Prefabricated vertical drains

Fold back the filter, secure the opening by tape and staple as shown :
Secure the opening by tape and staple as shown:

New roll Old roll New roll Old roll

Secured by staples Secured joint by tape Secured joint by tape Secured by staples

Figure 4.21 Details of splicing: (a) Colbond type of drain. (b) Mebra type of drain. (From Bo, M.W., Chu, J., Low, B.K. and Choa, V. (2003). Soil
Improvement: Prefabricated Vertical Drain Technique, Thomson Learning, Singapore.)

121
Table 4.4 Speciications used for different projects in different countries

122
Netherlands

Ground improvement
Description Unit Standard Case I Case II Singapore Thailand Hong Kong Malaysia Taiwan Australia Finland Greece
Tensile strength kN/10cm ASTM >0.5 >0.5 >1 (at >0.5 >2 >1 >1
(dry) D4595 10%)
Tensile strength >0.5 >0.5 >1 (at >2 >1 >1
(wet) 10%)
Elongation % 2–10 (at 0.5 kN) <30 <20 15-30
(at 1 kN) (yield)
Discharge capacity 10–6 m2/s ASTM >10 at >50 at >25 at 350 >16 at 200 >5 at 200 >6.3 at >10 at >100 at >10 >10
(straight) D4716 and 350 350 kPa 28 kPa, kPa 400 kPa, 300 300 kPa at
others kPa 30 kPa 30 days 7 days, i=1 kPa, 100
days days i = 1. i=1 kPa
Discharge capacity 10–6 m2/s >7.5 at >32.5 at >10 >6.3 at
(folded) 350 350 400 kPa
kPa kPa
Crushing strength kN/m3 500
Equivalent mm 50 65
diameter
Free surface ilter mm2/m 150,000
Tear strength N ASTM D4533 100 >300 >250 >380
Graph strength N ASTM D4632 >350
Puncture strength kN ASTM D4833 >200
Bursting strength kPa ASTM D3785 >900
Pore size O95 μm ASTM D4751 <160 <80 <75 <90 <120 <75 <90
Permeability of 10–5 m/s ASTM D4491 >5 >10 >1 >10 >17 >50
ilter
Case I: Stable layer less than 10 m thick; Case II: Unstable layer more than 10 m thick
Source: Modiied from Bo et al., 2003.
 Prefabricated vertical drains 123

Figure 4.22 Device to measure the permeability of ilter of PVD.

4.4.2.2 Permeability (or permittivity) of filter

The cross-plane (i.e., in a direction perpendicular to the surface of the ilter)
permeability, or the so-called permittivity, of the ilter is measured using a
constant head method. A permeameter specially made for this purpose
is shown in Figure 4.22. A single layer of ilter is used. ASTM Standard
D4491 99a (2009) is followed in conducting this test. In this case, the per-
meability measured under a constant head of 50 mm is reported.

4.4.2.3 Discharge capacity test

The discharge capacity of drain, qw, is normally measured for both straight
and deformed drains. This is because the discharge capacity of PVD
reduces after it has buckled in soil. Although the ASTM4716 (2008) is
often referred to, it should be pointed out that this standard is set for the
determination of the transmissivity of a geosynthetic, not speciically for
the discharge capacity of PVDs. Therefore, a standard method for measur-
ing the discharge capacity of vertical drain has not been established yet. As
such, various devices and methods have been developed for measuring the
discharge capacity of vertical drain (Hansbo, 1983; Kamon et al., 1984;
Guido and Ludewig, 1986; Suite et al., 1986; Broms et al., 1994; Bergado
et al., 1996; Chu et al., 2004). These methods can be generally classiied
into two categories. The irst is to embed the drain in soil and the second is
to warp the drain with membrane or other materials.
The testing devices suggested in the European Standard on Execution
of Special Geotechnical Works—Vertical Drainage (BS EN 15237, 2005)
124 Ground improvement

Water supply
Load

Head loss Overflow

Sample length 300 mm Foam Water collection

Figure 4.23 Testing apparatus #1 for discharge capacity test recommended by BS EN

15237 (2007).

for determining the discharge capacity according to EN ISO 12958

belongs to the irst category. The two devices are shown in Figures 4.23
and 4.24, respectively. The specimen in apparatus #1 is covered on both
sides by closed-cell foam rubber with a thickness of 10 mm. The mem-
brane in apparatus #2 is made of latex with a maximum thickness of
0.35 mm.
The studies of Lee and Kang (1996), Chu et al. (2004), and Karunaratne
(2011) indicated that soil should be used in the discharge capacity test as
the results obtained from the tests without soil tend to be substantially
larger than those with soil. This is mainly because of the greater indenta-
tion of ilter into the drain groove as shown in Figure 4.25, and the reduc-
tion in the water-discharging ability of the ilter when soil is in contact with
the ilter. For this reason, a method to embed a drain specimen in soil may
be a better alternative. Another problem related to the device suggested
in ASTM4716 or in BS EN 15237 (2007) are shown in Figures 4.23 and
4.24 is that the total head is not measured within the drain specimen. This
is necessary as it is often found out that the hydraulic gradient measured
between two points within the drain is different from that measured using
the head difference of the water reservoirs.
To improve the ASTM D4716 procedure, a new drain tester has been
developed and used for a number of land reclamation projects in Asia
(Broms et al., 1994; Chu and Choa, 1995; Chu et al., 2004). A cross-section
of the new drain tester is shown in Figure 4.26. It consists of a base, a hol-
low extension plate, and a top cap. In a test, a drain specimen of 100 mm
(or 300 mm) is placed in the base on top of a marine clay layer. The hollow
 Prefabricated vertical drains 125

Head loss 3

Manometer 8

Rubber membrane 9

Flow meter 10

Flow direction 1
Sample length

Figure 4.24 Testing apparatus #2 for discharge capacity test recommended by BS EN

15237 (2007).

extension plate is used to anchor the drain and to allow another layer of
marine clay to be put on top of the drain. The top cap is then placed and
the screws are tightened. Square shaped O-rings are used to seal the tester.
The vertical pressure is applied via an oedometer loading frame. This test
is simple and can be easily conducted in a site laboratory. The results are

Soil

Filter

Core

Figure 4.25 Reduction of low area caused by the deformation of ilter.

126 Ground improvement

Allen screw

Top cap

Hollow
‘O’ Rings
extension plate
Marine clay
Drain 100mm × 100mm

Marine clay
Valve Valve

Base

Figure 4.26 A cross-section of the straight drain tester.

easily reproducible and therefore can be counter checked easily. This test
also provides a mean to compare and benchmark the discharge capacity
values of different drains. The discharge capacity measured using this new
drain tester is lower than that measured by the ASTM4716 method. It is
safer to use the smaller value measured to compare with the speciication.
A buckled drain tester, as shown in Figure 4.27, is also used to measure
the discharge capacity of the deformed drain (Chu and Choa, 1995). A
buckled sample is shown in Figure 4.28. In conducting this test, it is not
necessary to wait until the soil in the drain tester has achieved a high degree
of consolidation as the dissipation of excess pore pressure should not affect
the discharge capacity measurement. However, it is troublesome to carry
out a discharge capacity test using a tester shown in Figure 4.25. An alter-
native is a method suggested in BS EN 15237 (2007) using an apparatus
shown in Figure 4.29 in which a drain specimen inside a rubber membrane
is bent in a way shown in Figure 4.29. Another method is to use a device
shown in Figure 4.30 where a drain specimen is bent to a 30-degree angle
(Figure 4.30b) using two pieces of water impervious foam (Bo et al., 2003).
It should be mentioned that as a standard, the discharge capacity should
be reported as the value measured at 20°C. If the discharge capacity test
is not conducted at this temperature, a conversion in the same way as for
permeability test should be made.
It should be pointed out that the discharge capacity also reduces with
time. This might be due to the creep of the drain materials under pressure.
 Prefabricated vertical drains 127

Water discharge

To piezometer
board

Air vent Air pressure

Marine clay

Vertical drain

Geotextile

Pressure transducer 2 Pressure transducer

Water inlet
To piezometer
board

Figure 4.27 Buckling drain tester.

Figure 4.28 Deformed drain after it is tested in the buckled drain tester.
128 Ground improvement

Rod

50
Drain specimen

Guide rod 30

Figure 4.29 Device for measuring the discharge capacity of a bent drain recommended
by BS EN 15237 (2007).

The reduction in the discharge capacity over four weeks can be as high as
60% (Chu and Choa, 1995). For this reason, a creep factor is introduced to
account for the creep effect. Although it is too time consuming to measure
the long-term discharge capacity, some long-time discharge capacity tests
should be conducted to assess the rate of reduction in discharge capacity
with time or verify the creep factor adopted.
It is generally observed that the discharge capacity reduces with hydraulic
gradient (Kamon et al., 1984; Broms et al., 1994; Park and Miura, 1998).
As such, when the value of discharge capacity is reported, the hydraulic
gradient, i, at which the discharge capacity is measured should be stated.
For practical purposes, the discharge capacity should be measured at a
hydraulic gradient comparable to the ield conditions. However, the in-situ
hydraulic gradient is dificult to estimate. Reports on the ield hydraulic
gradient are also rare, except one case reported by Nakanado et al. (1992)
in which the in-situ hydraulic gradient was estimated to be in the range of
0.03–0.8. From the testing point of view, Akagi (1994) pointed out that
when the hydraulic gradient is higher than 0.5, the low inside the vertical
drain may not be laminar anymore. He suggested the discharge capacity

(a) (b)

Rod
Drain specimen

Guide rod

Figure 4.30 Device for measuring the discharge capacity of deformed drain.
 Prefabricated vertical drains 129

160
Straight

140 Buckled
Discharge capacity (10–6 m3/s)

Hydraulic gradient = 0.5

120

100

80

60

40

20
0 50 100 150 200 250 300 350 400
Pressure applied (kPa)

Figure 4.31 Typical PVD discharge capacity test results using the straight and buckled
drain testers.

to be measured at a hydraulic value ranging from 0.2–0.5. After analysing

the low behaviour in the drain under different hydraulic gradients, Park
and Miura (1998) also suggested that a hydraulic gradient ranging from
0.2–0.5 be used. The data presented in Wang and Chen (1996) also indi-
cate that a steady low can be dificult to achieve for vertical drain when i >
0.5. Based on these studies, an i ≤ 0.5 should be used for discharge capac-
ity measurement. BS EN 15237 (2007) suggests the discharge capacity be
measured at a hydraulic gradient of i = 0.1. However, the testing errors
involved in the measurement can be higher when the hydraulic gradient
is small. Therefore i = 0.5 appears to be the most suitable value. In our
method, the discharge capacity of PVD was determined by measuring the
discharge capacity at different i values ranging from 0.1 to 1 to establish the
relationship between qw and i under each pressure irst, and then using this
relationship to determine the discharge capacity at i = 0.5. The qw measured
in this way has been used in Inequality (4.20), which has been shown to be
compatible with ield measurements, as discussed in the early section. Some
typical results of discharge capacity tests for a PVD in both straight and
buckled conditions are shown in Figure 4.31.

4.4.2.4 Tensile strength tests

ASTM D4632 (2008) is often speciied as the method to measure the ten-
sile strength of PVD or the ilter of PVD, in which the pull rate is 300 ±
10 mm/min. A tensile strength testing machine that provides such a pull
130 Ground improvement

rate is normally too expensive for a site laboratory to equip. On the other
hand, a compression machine used for triaxial tests is commonly available
in a geotechnical laboratory. Therefore, a method that uses a modiied
compression machine to conduct tensile strength tests for vertical drains
has been developed (Chu and Choa, 1995). The only shortcoming of using
a compression machine is that it does not provide a pulling rate as high
as 300 mm/min. However, this will not be a problem for the following
reasons—irst, the drain installation speed is as high as 25,000 mm/min.
As even 300 mm/min is far too low, whether a pulling rate is 300 mm/min
or lower does not make much difference. Second, the slower the rate, the
smaller the tensile strength measured, and the use of a slower pulling rate
will result in a more conservative measurement. This is good for quality-
control purposes.
When a compression machine is used to conduct tensile strength tests,
a pair of clamps designed according to ASTM D4632 need to be used to
clamp the drain specimen to the machine. A drain specimen of 200 mm in
gauged length is used. The test setup is shown in Figure 4.32. The load was
applied under constant-rate-of-extension (CRE). For wet conditions, the
specimen, either the entire drain or the ilter, was immersed in water for 48
hours before testing. Some typical tensile strength tests for two different
types of PVDs are shown in Figure 4.33.

Figure 4.32 Tensile strength test for PVD or the ilter.

 Prefabricated vertical drains 131

2.4
Dry
Drain 1
2
Wet

1.6
Tensile stress (kN)

Dry
Drain 2

1.2 Wet

0.8

0.4

0
0 10 20 30 40
Tensile strain (%)

Figure 4.33 Tensile stress versus tensile strain curves measured for 2 typical drains.

4.4.3 Measurement of penetration depth

The effect of soil improvement using PVD is greatly affected by the instal-
lation depth. If a full penetration of PVDs is assumed in the design and yet
the PVDs are not installed to the entire depth of soft clay, the predicted rate
of consolidation will be incorrect. Therefore, it is important to measure
the real penetration length of the PVD installed on site. Another reason for
measuring the installation depth of PVD is to gain a more speciic knowl-
edge of the depth of soft soil at the PVD installation locations. The instal-
lation depth of PVD is normally speciied by the designer. However, when
erratic soil proiles are encountered, contractors are allowed to terminate
the PVD only when the stiff or hard formation below the soft soil forma-
tion is encountered, which can be gauged based on the efforts required to
penetrate the mandrel. In this case, the thicknesses of the soft soil layer at
different PVD installation points can be known.
At the present, the following three methods have been adopted in mea-
suring the penetration depth of PVD as described by Bo et al. (2003): (1)
using a meter on the mast; (2) using a dial gauge; and (3) using an automatic
digital counter. However, all the three methods measure only the length of
the PVDs that pass through the point where the counter or the dial gauge
is located, not the real length of the PVDs that has been installed into the
soft clay. For this reason, none of the three methods can provide a direct
measurement of the PVD installed in the ground. Therefore, none of the
three methods is suitable to be used for independent checking or auditing
purposes. Without measuring the penetration depth of the PVD directly, it
will be impossible to check whether there is any mistake or even cheating
132 Ground improvement

in the PVD installation records. There were cases where PVDs were not
installed deliberately to the required depths. Therefore, a method that can
directly measure the installation depth of PVD is also required. In the fol-
lowing, three new methods that can measure the penetration depth of PVDs
directly after the PVD has been installed are introduced. These three new
methods are digitised PVD, PVD with two wires, and PVD with one wire.

4.4.3.1 Digitised PVD

The irst method is to print a meter scale on the surface of PVD at an inter-
val of 20 or 25 cm so its linear length can be read directly. An example
is shown in Figure 4.34. The meter scale can be printed onto the surface
of PVD automatically using the computer scale-spattering digital metering
technology. When the PVDs are installed by following a given sequence,
the differences in the meters printed on top of the current and the last PVD
installed will be the installation length of the current PVD. The total length
of PVDs used can also be calculated easily. This method is simple and
incurs almost no extra cost. However, it has the following shortcomings.
The numbers printed on the ilter of PVD become illegible when the PVDs
are stored or exposed outside for too long. Also, the readings are affected
by the expansion or contraction of PVDs due to temperature variation or
wetting. PVDs may be stretched during installation and the scale printed
may not be accurate. Furthermore, it is still not a direct measurement as the
length of the PVDs is not measured directly.

Figure 4.34 PVD with scale printed on it for PVD penetration depth measurement.
(From Liu, H.L., Chu, J. and Ren, Z.Y. (2009). Geotextiles and Geomembranes,
Vol. 27, No. 6, pp. 493–496.)
 Prefabricated vertical drains 133

4.4.3.2 PVD with two wires

The second method is to embed two shielded thin copper wires in the PVD,
as shown in Figure 4.35. This method has been patented (Ren, 2004). The
length of the PVD can be calculated by measuring the resistance of the
wires. The two wires are embedded along the overlapping joint of the ilter.
Before installation, the two wires at the bottom end of the PVD need to be
connected together. At the top end of the PVD, the wires are connected to
a meter to measure the electrical resistance of the two wires as one loop. A
readout unit as shown in Figure 4.36 has been specially designed for this
purpose. This readout unit can measure the resistance of the wires, convert
it directly into length, and display and store the readings. This is probably
the most direct and reliable method available so far. However, this method
also has some shortcomings. Firstly, it incurs extra costs to the PVDs to use
two wires. Secondly, the connection of two wires before each installation of
PVD is troublesome. Furthermore, the connection has to be done properly
or the method will not work. Disputes may rise sometimes on whether the
PVD is not installed properly or simply because the wires at the end of the
PVD are not connected properly. For this reason, a standard procedure
should be adopted. It is suggested to use a minimum connection length of
20 mm. The insulation at the connection must be removed by burning or
scratching. The connected portion should be put back into the ilter.

Figure 4.35 PVD with two copper wires embedded for PVD penetration depth. (From
Liu, H.L., Chu, J. and Ren, Z.Y. (2009). Geotextiles and Geomembranes, Vol. 27,
No. 6, pp. 493–496.)
134 Ground improvement

Figure 4.36 Readout unit for the measurement of PVD penetration depth. (From Liu,
H.L., Chu, J. and Ren, Z.Y. (2009). Geotextiles and Geomembranes, Vol. 27,
No. 6, pp. 493–496.)

4.4.3.3 PVD with one wire

To overcome the problems associated with the two-wire PVD method, a
third method has been developed. This method is similar to the two-wire
PVD method, but uses only a single wire. This thin copper wire is embedded
along the overlapping joint of the ilter as shown in Figure 4.37. This method
is based on the principle of microwave impedance measurement. When PVD
is installed into the ground, the wire in the PVD and another wire connect-
ing to the ground as provided by the readout unit form a two-wire system.
Some ield veriication of this method is given in Liu et al. (2009).

4.4.4 Field instrumentation and evaluation of performance

A preloading plus PVDs soil improvement project is usually carried out until
the required degree of consolidation is obtained. Degree of consolidation is
an important parameter in evaluating the effectiveness of soil improvement.
It is also often used as a design speciication in a soil improvement contract.
Assessment of the degree of consolidation of the soil therefore becomes one
of the most important tasks for construction control. One of the most suit-
able methods for assessing the degree of consolidation of soil is by means of
ield instrumentation using settlement or pore water pressure data. For this
reason, ield instrumentation is normally required to monitor settlements and
 Prefabricated vertical drains 135

Figure 4.37 PVD with one copper wire embedded for PVD penetration depth mea-
surement. (From Liu, H.L., Chu, J. and Ren, Z.Y. (2009). Geotextiles and
Geomembranes, Vol. 27, No. 6, pp. 493–496.)

pore water pressures at different elevations as well as groundwater tables,

lateral displacement, and earth pressure. One example, for the Changi rec-
lamation project in Singapore, is shown in Figure 4.38. It can be seen that
in each soil layer, at least three pore pressure transducers should be used to
construct the pore pressure distribution versus depth proile. Some of the
commonly used instruments are described in Bo et al. (2003). A case study is
presented and some instrumentation issues are discussed in Arulrajah et al.
(2009). For projects using PVDs, in particular those that use vacuum pre-
loading, it is highly desirable to measure the pore water pressure inside the
PVD. A piezometer can be placed inside a PVD, as shown in Figure 4.39
before PVD installation for this purpose. Some typical settlement and pore
water pressure monitoring data are shown in Figure 4.40.
The degree of consolidation is normally calculated as the ratio of the
current settlement to the ultimate settlement. However, for a soil improve-
ment project, the ultimate settlement is unknown and has to be predicted.
Although consolidation settlement can be estimated based on laboratory
oedometer tests, the prediction by this method is normally not very reli-
able. Methods to estimate the ultimate settlement based on ield settlement
monitoring data are also proposed. Among them, the Asaoka (1978) and
hyperbolic (Sridharan and Rao, 1981) methods are commonly used.
In Asaoka’s method, a series of settlement data (S1,..., Si−1, Si, Si+1,...SN)
which are observed at constant time intervals are plotted in a Sn versus Sn−1
plot (n = 1,..., N). The ultimate settlement, Sult, is taken as the intersecting
point of the curve itting line with the 45° line (Asaoka, 1978), as illustrated
in Figure 4.41. However, Sult obtained from Asaoka’s method is affected by
 136
Ground improvement
FT–2: Prior to reclamation
FT–9: After ground improvement FT–8 (A2S–6): After ground improvement
PB–88
PB–39
PB–87
DS SP–04
PP WS–04 DS SP 095 PP WS–20
+10 mCD
Surcharge
+4 mCD

Sandfill

–4.10 DS 093 –6.30

–7.44 PP 20 PP 076
Upper marine clay
–10.80 DS 106 –9.50 PP 077
–14.44 DS 012–11.44 PP 21
–14.80 DS 107 –12.70 PP 078
–17.24 DS 013–15.94 PP 22
–19.44 DS 014 –17.40
–18 44 PP 23 PP 079
Intermediate –22.00 DS 108
–22.44 PP 59 –22.44 PP 080
stiff clay
–28.74 DS 015 –26.44 PP 24 –27.10 DS 109
–27.50 PP 081
–33.84 –31.44 PP 25 Lower marine clay –32.00 PP 082
DS 016
–34.10 DS 110

Old alluvium

Figure 4.38 Typical details of onshore and adjacent offshore ield instrumentation clusters. SP = Settlement plate, DS = Deep settlement gauge,
PP = Pneumatic piezometer, WS = Water stand-pipe. (From Arulrajah, A., Bo, M.W. and Chu, J. (2009). Instrumentation at the
Changi land reclamation project, Singapore. Geotechnical Engineering, Proceedings of the Institution of Civil Engineers, Vol. 162, No. 1,
pp. 33–40.)
 Prefabricated vertical drains 137

Figure 4.39 Putting piezometer inside a PVD for pore water pressure measurement.

the time interval used (Matyas and Rothenburg, 1996; Bo et al., 1999). In the
hyperbolic method, settlement data are plotted as time/settlement versus time
curve (Sridharan and Rao, 1981). The Sult is estimated as the inverse of the
linear slope of the plot. However, Sult obtained from this method is affected
by the degree of consolidation achieved. The higher the degree of consoli-
dation that the soil has attained, the smaller the Sult obtained as observed
by Matyas and Rothenburg (1996), Bo et al. (1999), and Goi (2004). The
uncertainties involved in the ultimate settlement calculation will affect the
estimation of the degree of consolidation. As a result, different degrees of
consolidation are obtained using different methods. As an alternative, pore
water pressure data can be used to assess the degree of consolidation.
Once the pore water pressures at different depths are measured dur-
ing preloading, the initial and inal pore water pressure distributions with
depth can be plotted (Chu et al., 2000). For generality, a combined ill
surcharge and vacuum load case is considered. The typical pore water pres-
sure distribution proiles for a combined vacuum and ill surcharge loading
case are shown schematically in Figure 4.42. Using this proile, the average
degree of consolidation, Uavg, can be calculated as:

Uavg = 1 −
∫ [u (z) − u (z)] dz
t s
(4.24)
∫ [u (z) − u (z)] dz
0 s

and
us(z) = γwz − σ, kPa.
 138
Elapsed time (day)

Ground improvement
9 109 209 309 409 509 609 709 809 909 1009
Soil profile Soil 10
Ref. Borehole DC-013 instrument

Surcharge (mCD)
E.L. Dept Soil 8
(mCD) (m) type E.L. (mCD)
0 Sand SP-082(3.51)
4 6
4
DS-095(–8.4)
–8.4 12.4 Clay 2
PP-110(–10)

DS-096(–14) 0

0 DS-104
PP-112(–18) DS-101 DS-100
DS-097(–20) 0.5 DS-99
Settlement (cm)

DS-098(–24) DS-98
PP-114(–25.8) 1 DS-97
DS-099(–27.8)
1.5 DS-96
–28.02 32.02 Sand
–29.88 33.88 Clay 2 DS - Deep Settlement Gauge
PP-116(–35) DS-95
DS-101(–36) 2.5 SP - Surface Settlement Plate
SP-82
–38.93 42.93 Silt 3
PP-119(–47)
20
PP - Pneumatic Piezometer
PP El. (mCD)

16 PP-119 PP-116 PP-114

PP-112 PP-110 Water level
12
DS-104(–57.4) 8
–57.4 61.4 Sand
4
–63 67 0
0 100 200 300 400 500 600 700 800 900 1000 1100
Elapsed time (day)

Figure 4.40 Typical settlement and pore water pressure data obtained from the third pilot test. (From Chu, J., Bo, M.W. and Choa, V. (2006).
Geotextiles and Geomembranes, Vol. 24, No. 6, pp. 339–348.)
 Prefabricated vertical drains 139

Sult

Settlement at time t, Sn Sn

S3

S2

45°
S1 S2 ....... Sn-1
Settlement at time t-1, Sn-1

Figure 4.41 Schematic illustration of Asaoka’s method.

Hydrostatic pore
water pressure
line, uw (z)

ut (z)
Depth, z (m)

u0 (z)

Suction line, us (z)

S ∆σv
Pore water pressure (kPa)

Figure 4.42 Schematic illustration of pore water pressure distributions versus depth
under combined surcharge and vacuum load.
140 Ground improvement

In Equation 4.24, u 0 (z) = the initial pore water pressure at depth z;

ut (z) = the pore water pressure at depth z at time t; us(z) is the suction line,
γw = unit weight of water, and s = suction applied. The value of s is nor-
mally assumed to be 80 kPa. The integral in the numerator in Equation
4.24 is the area between the curve ut(z) and the suction line u s(z), and
the integral in the denominator the area between the curve u 0(z) and the
suction line u s(z).
The method shown in Equation 4.24 has the following advantages over
the method using settlement data:

(1) The degree of consolidation calculated using Equation 4.24 relies

only on ield pore water pressure data, whereas when calculating the
degree of consolidation using settlement data, the ultimate settlement
has to be predicted.
(2) Not only the inal degree of consolidation, but also the degree of con-
solidation at any time can be calculated using Equation 4.24, as ut(z)
represents the pore water pressure at any time, t.
(3) For consolidation involving multiple layers, Equation 4.24 can be
applied to any single layer to calculate the degree of consolidation
achieved in a particular layer. In this case, the upper and lower limits
of the integrals in Equation 4.24 are set to be the top and bottom
of that soil layer. However, it is not easy to calculate the degree of
consolidation for each layer for multilayer soils using settlement, as
the settlement of each layer may not be monitored directly and the
ultimate settlement of each layer has to be predicted, too.

With the use of PVD, the pore pressure transducers are installed within
half of the drains’ spacing distance to a PVD. This distance may vary
depending on the verticality of the PVD during installation and the
subsequent deformation of the PVD. One concern is that the random
uncertainties of the distance between the pore pressure transducers and
the PVDs will affect the pore pressure distribution proile, as shown in
Figure 4.40. This is true only when the depth of PVD is relatively short,
say, less than 10 m. This is because when a random variable varies over a
long distance, the overall effect of the random variation over the entire dis-
tance reduced greatly due to a statistical property called spatial variance
reduction. This explains why the method illustrated in Equation 4.21
has worked well for a number of projects (Chu et al., 2000; Chu and
Yan, 2005; Yan and Chu, 2003; 2007; Chu et al., 2009). More details
and comparisons are also made in the case histories published in these
references.
Degree of consolidation may also be estimated using the undrained shear
strength distribution proile. One example will be shown in the Case Study
in Section 4.7.
 Prefabricated vertical drains 141

It should be mentioned that any method can only be as reliable as the

ield monitoring data. Any uncertainties involved in the ield pore water
pressure and settlement measurements will inevitably affect the degree
of consolidation estimation. Furthermore, when pore water pressures are
measured at only a limited number of points, the spatial pore water pres-
sure distribution cannot be constructed. In this case, the pore water pres-
sure distribution proile established for one section has to be assumed to
be the same as that at other sections. This may not be the case, although
with the use of PVDs the pore water pressure distributions tend to even out.
For the degree of consolidation estimated based on settlement, there is one
more source of uncertainty, that is, the uncertainties involved in the ulti-
mate settlement prediction. In view of the various uncertainties involved
in the degree of consolidation calculation, it is recommended to estimate
the degree of consolidation using both settlement and pore water pressure
data. Even if the degree of consolidation is to be calculated using settle-
ment data, the pore water pressure distribution proile provides a way to
visualise whether the pore water pressure dissipation is consistent with the
degree of consolidation calculated based on settlement. If the differences
between the two measurements are too large and the difference cannot be
explained, the results should then be examined before they are accepted.
For contracting purposes, it will be necessary to specify clearly whether the
degree of consolidation should be evaluated based on settlement or pore
water pressure or both to avoid future dispute.

4.5 DESIGN CODES, STANDARDISATION,

AND SPECIFICATIONS

As discussed in the preceding sections, a number of ASTM D-series stan-

dards have been used as the standards for vertical drain testing. However,
some of those ASTM standards are not speciically written for vertical
drains. Therefore, the testing procedures stipulated in these standards may
not be the most suitable methods. Some of the ASTM standards can be
compiled using different testing systems. This has been the reason why
there are so many different testing methods proposed. As the performance
of vertical drains can be affected considerably by the quality of the drains
used and the control in the construction procedures, it would be highly
desirable to set up some regulations or codes of practice to govern the selec-
tion of vertical drain and to regulate the construction activities. Several
design codes or standards have been developed. These include the European
Standard on Execution of Special Geotechnical Works—Vertical Drainage
(BS EN 15237, 2007), the Chinese design code JTJ/T256-96 (1996) that
controls the practice for installation of PVDs and JTJ/T257-96 (1996) that
stipulates the quality inspection standard for PVDs, and the Australian
142 Ground improvement

Standards for the execution of PVDs, AS8700 (2011). The use of these
codes and standards is important in maintaining the quality standards of
soil improvement works.
The Chinese Quality Inspection Standard for PVDs (JTJ/T257-96)
requires every batch of PVDs and every 200,000 m within the same batch
to be sampled and tested for quality-control purposes. What is interest-
ing is that JTJ/T257-96 speciies the thickness of PVDs to vary according
to the depth of installation as well as the discharge capacity and tensile
strength (see Table 4.5). Although no reasons were given for the speciic
values used, the use of a thicker and thus larger discharge capacity PVD
is in line with the requirement expressed in Inequality (4.20). The other
recommended speciications for PVDs are also given in Table 4.5.
The European Standard BS EN 15237 includes the application of PVDs
and sand drains and deals with requirements to be placed on design,
drain material, and installation methods. For the material properties of
PVD, the main properties required and their testing methods are listed in
Table 4.6. All the PVD properties used in Table 4.6 have been explained in
this chapter except for the velocity index of ilter vh50 and durability. The
‘velocity index of ilter’ sounds new, but is merely another way to measure
ilter permeability in Europe. It deines the iltration velocity corresponding

Table 4.5 Minimum thickness and other speciications for PVD as speciied by the
Chinese Code JTJ/T257-96
(a) Minimum thickness
Type L < 15 m L < 25 m L < 35 m Stud type
Thickness (mm) > 3.5 > 4.0 > 4.5 >6
(b) Other speciications
Description Unit L < 15m L < 25m L < 35m Testing conditions
Discharge cm /s
3 15 25 40 Under pressure of
capacity (m3/yr) (670) (1,115) (1,784) 300 kPa
Permeability of cm/s 5 × 10–4 After the sample is
ilter immersed in water for
24 h
Pore opening of μm < 75 O95
ilter
Tensile strength kN/10 > 1.0 > 1.3 > 1.5 At 10% elongation
of PVD cm
Tensile strength kN/cm > 15 > 25 > 30 At 10% elongation
of ilter (dry)
Tensile strength kN/cm > 10 > 20 > 25 At 10% elongation. Sample
of ilter (wet) immersed in water for
24 h.
 Prefabricated vertical drains 143

Table 4.6 Requirements of PVDs properties in BS EN 15237 (2007)

Properties Requirement Testing Method
Tensile strength > 1.5 kN at failure EN ISO 10319
Elongation at max. tensile force elongation ≥ 2% at failure; EN ISO 10319
elongation ≤ 10% at a
tensile force of 0.5 kN
Tensile strength of ilter > 3 kN/m or 6 kN/m for EN ISO 10319
PVD longer than 25 m.
Tensile strength of seams and joints > 1 kN/m EN ISO 10321
Velocity index of ilter (vh50) > 1 mm/s EN ISO 11058
Characteristic opening size of ilter (O90) < 80 μm; and EN ISO 12956
< (1.5 to 2.8)d50 of soil
Discharge capacity of the drain See Figure 4.43 EN ISO 12958
Durability PVDs to be covered on the EN 13252
same day

10

0.5
Permeability of soil, m/year

0.1

0.05 2000
Discharge capacity, m3/year

1000

0.01 500

0.005 200

100

0.001 50
5 10 20 30 40 50
Depth of installation, m

Figure 4.43 Required discharge capacity qw as a function of permeability of soil and

depth of drain installation shown in BS EN 15237 (2007)
144 Ground improvement

to a head loss of 50 mm across the specimen (CEN, 1998). Durability

is a general requirement for geotextile- and geomembrane-related prod-
ucts. It is stated in EN 13252 that all the geotextile- and geomembrane-
related products shall pass the accelerated weathering test according to EN
12224, unless they are to be covered on the day of installation. Therefore,
if we ensure the PVD is to be installed within one day after it is taken out
from the store, durability will not be a problem. It is noted that the dis-
charge capacity is not speciied as a ixed value, but a function of depth of
installation and the permeability of soil as shown from Figure 4.43. It can
be seen that the range of values is similar to the Chinese Code JTJ/T257-
96. In the past, many PVD speciications used a ixed discharge capacity
(some examples are given in Bo et al., 2003). In this respect, both the
Chinese Code JTJ/T257-96 (1996) and BS EN 15237 (2007) have set a
better design standard of PVDs.
BS EN 15237 has also speciied the frequency for conducting quality-
control tests as shown in Table 4.7. It is necessary to carry out quality-
control tests at a certain frequency, particularly when a huge quantity of
PVDs is used. Variations between the quality of PVDs were observed in
some past projects.

Table 4.7 Proposed testing frequency for fabrication control

Proposed test Required
Property frequency standard

Filter:
Thickness 25,000 m2 EN 9863-1
Mass per unit area 25,000 m2 EN 9864
Pore size 200,000 m2 EN 12956
Velocity index 200,000 m2 EN 11058
Tensile strength in the longitudinal direction 200,000 m2 EN 10319
Tensile strength in the cross direction 200,000 m2 EN 10319

Drain composite:
Width and thickness 25,000 m EN 9863-1
Mass per unit length 25,000 m –
Tensile strength in the longitudinal direction 100,000 m EN 10319
Elongation at maximum tensile force 100,000 m EN 10319
Discharge capacity straight 500,000 m BS EN 15237
Discharge capacity buckled 500,000 m BS EN 15237
Tensile strength of ilter seam 100,000 m EN 10321
Durability 500,000 m EN 13252
Source: BS EN 15237. (2007). European Standard on Execution of Special Geotechnical
Works—Vertical Drainage. European Standard.
 Prefabricated vertical drains 145

4.6 PVD FOR VACUUM PRELOADING

When PVDs are used together with vacuum preloading for soil improve-
ment, some special arrangements may be required. The PVDs in this case
will not only discharge water, but also transmit vacuum pressure as shown
in Figure 4.44. Therefore, the PVDs used for a vacuum preloading project
should possess better quality than normal. Sometimes, prefabricated circu-
lar drains (see Figure 4.10) are also used. More detailed description of the
vacuum preloading system shown in Figure 4.44 can be found in Chu and
Yan (2005b).
The vacuum preloading system shown in Figure 4.44 works well when
the low permeability compressible soil extends all the way to the ground
surface so that membranes can be used to seal the top surface for vacuum
pressure application. However, when there is a relatively thick layer of per-
meable soil on top, a cut-off wall extending all the way to the bottom of

A A

1 2 3 4 5 6 7 8

8 7 4 10 11

A–A

Figure 4.44 Schematic arrangement of vacuum preloading system. 1, drains; 2, ilter

piping; 3, revetment; 4, water outlet; 5, valve; 6, vacuum gauge; 7, jet pump;
8, centrifugal gauge; 9, trench; 10, horizontal piping; 11, sealing membrane.
(From Chu, J., Yan, S.W. and Yang, H. (2000). Géotechnique, Vol. 50, No. 6,
pp. 625–632.)
146 Ground improvement

Plastic pipe

Sand layer

PVD

Clay
layer

(a) (b)

Figure 4.45 BeauDrain vacuum preloading system. (a) Concept. (Courtesy of Cofra, Holland.)
(b) Direct connection of PVD with plastic pipe for vacuum application.

the permeable soil layer will have to be installed around the whole soil
improvement site. This can be too expensive. An alternative is to connect
each individual PVD to a plastic pipe as shown in Figure 4.45. The plastic
pipe can be installed together with PVD into the ground and going through
the entire permeable soil layer as shown in Figure 4.45a. In this way, the
need to use cut-off wall and membrane for creating an airtight seal is no
longer required. This so-called BeauDrain technique has been developed by
Cofra Holland (Kolff et al., 2004) and used for a number of soil improve-
ment projects (Seah, 2006; Saowapakbiboon et al., 2008). However, this
method has its own limitations. First, it requires the soil proile at the PVD
installation points to be known precisely. This can be dificult sometimes.
Second, the length of each PVD and the plastic pipe connected has to be
determined and preconnected beforehand. Third, for installation, each
PVD with the plastic pipe has to be pulled through a mandrel with the
plastic pipe on top and then installed into the ground using a PVD instal-
lation machine, see Figure 4.46. The plastic pipes are then connected to a
main vacuum line which is linked to a vacuum pump, see Figure 4.47.
 Prefabricated vertical drains 147

Clay
layer

Figure 4.46 Installation of PVDs with plastic pipe connected.

Figure 4.47 Connection of plastic pipes to a vacuum pump.

4.7 CASE STUDIES

4.7.1 Reclamation and soil improvement

for a slurry pond in Singapore
PVDs have been used in many soil improvement and land reclamation
projects. However, the use of PVDs for the improvement of ultra-soft
soil; that is, soil with no or little shear strength are still not common.
One such a project was carried out in Singapore to reclaim a slurry pond
of about 180 hectares as part of the Changi East Reclamation Projects
(Choa et al., 2001, Bo et al., 2005). This slurry pond was created by
148 Ground improvement

dredging seabed to an elevation of −22 mCD (Chart Datum) for the stor-
age of silt and clay washings from other sand-quarrying activities. The
thickness of the ultra-soft slurry varied from 1–20 m with an average
value of 15 m. The grain size distribution curves indicate a ines content
in the range of 70%–93%. The upper bound of the mean grain size D50
was 0.024 mm, but mostly in the range smaller than 0.001 mm and D85
was in the range of 0.004–0.02 mm. The water content of the slurry was
mainly in the range of 140%–180%. The bulk unit weight of the slurry
ranged mainly from 11–13 kN/m 3. As the slurry was deposited recently
with little consolidation, it was ultra-soft and highly compressible.
Based on the properties of the slurry and the depth of installation, the
properties of the PVDs were chosen as shown in Table 4.8. The analysis
for the selection of PVD is detailed in Chu et al. (2006).
The procedure adopted for the reclamation of the slurry pond was plac-
ing a sand-capping layer on top of the slurry before PVDs could be installed
and used for the consolidation of the slurry. As the slurry had essentially
no strength, the land reclamation work was carried out by spreading thin
layers of sand using a specially designed sand spreader (Chu et al., 2009).
To ensure the stability of the ill, small lifts of 20 cm were used in the irst
phase of the spreading. This phase of sand spreading took about 13 months
including the waiting time between the lifts. There was a slurry burst when
the ill reached an elevation between –1 and +2 mCD (Chu et al., 2009).
Otherwise, the reclamation using sand spreading was workable. After the
ill was exposed above the water level, PVDs were installed with 2 × 2 m
square spacing. The surcharge was then placed to +6 mCD. The settlement
of the ill was monitored. After about 1.5 m of settlement had occurred,
a second round of PVDs with the same 2 × 2 m spacing was installed.
During the installation of PVDs, slurry was observed to come out through
the annulus of the mandrel, as shown in Figure 4.48. This is indicative
that the pore water pressure in the ground was still high and the installa-
tion process itself helped in the dissipation of pore water pressure and thus
the consolidation process. This is another advantage of the two-stage PVD
installation method.
Settlement gauges and pore pressure transducers were installed in the
ultra-soft clay layer to monitor the consolidation process of the ultra-soft
soil. The typical arrangement of instrument layout and proile are shown in
Figure 4.49. The surcharge history and settlement and excess pore pressure

Table 4.8 Properties of the PVD used

Discharge capacity of Discharge capacity of buckled
Permeability of straight drain under drain (under 25% strain) Apparent opening
the ilter 100 kPa pressure under 100 kPa pressure Size O95
2.6 × 10–4 m/s 70 × 10–6 m3/s 20 × 10–6 m3/s Less than 0.07 mm
 Prefabricated vertical drains 149

Figure 4.48 Extrusion of viscous pore water and mud during PVD installation. (From
Chu, J., Bo, M.W. and Choa, V. (2006). Geotextiles and Geomembranes, Vol. 24,
No. 6, pp. 339–348.)

(a) (b) SP 506 WS 110

+6.0 mCD

+4.0 mCD
+3.50
PP 467
DS 504 PZ 049
DS 505 PP 468
SP 506 PZ D48
DS 503 PP 466 Sand
DS 507 PP 047
DS 511 WS 110

PBS 5
–5.30 DS 503
Very soft slurry soil
–8.00 DS 504 PP 446 PZ 047

3m –10.00 DS 505 PP 487 PZ 048

–12.00 DS 511 PP 468 PZ 049

–13.0 mCD
3m

–14.30 DS 507
Sand

Figure 4.49 Arrangement of instrumentation. (a) Plan view. (b) Elevation view. (From
Chu, J., Bo, M.W. and Choa, V. (2006). Geotextiles and Geomembranes, Vol. 24,
No. 6, pp. 339–348.)
150 Ground improvement

versus time curves are shown in Figure 4.50. As indicated in Figure 4.50a,
the second round of PVD installation was carried out about 80 days after
the irst round of PVD installation, where the top of the slurry had settled
nearly 1.5 m. It should be noted that the surcharge load was reducing with
time as the ill used for surcharge was gradually submerged into water due
to the settlement of the ground. As shown in Figure 4.50b, the slurry had
settled for about 2.7 m in 500 days.
The excess pore pressures versus time curves measured by piezometers
PZ047, PZ048 and PZ049 (see Figure 4.49) are shown in Figure 4.50c.
The piezometers were installed after the irst PVD installation. A quick
increase in pore pressure was observed after 90 days at all three locations.
This was caused by the installation of the 2nd round of PVDs. The excess
pore pressure dissipations as measured by PZ047 at −8 mCD and PZ048 at
−10 mCD were slow despite of the occurrence of large settlement. The lack
of pore pressure dissipation may signify a sedimentation and self-weight
consolidation stage prior to consolidation under additional ill. During this
stage, slurry was transforming from a liquid to a solid state in which water
was dissipating, but the soil particles did not have suficient contacts to
allow the soil skeleton to take up external load. The Mandel–Cryer effect
and non-uniform consolidation of soil around the PVD were thought to be
the other reasons accounting for the lack of pore water dissipation (Chu
et al., 2006). The pore pressure dissipation measured by PZ-49 at −12
mCD was relatively quick. This was because the soil at this elevation was
near the silty sand layer below it.
Based on the pore water pressure measurements shown in Figure 4.48c,
the pore water pressure versus depth proile can be plotted in Figure 4.51.
The initial excess pore water pressure, which had the same magnitude as
the surcharge and the hydrostatic pore water pressure line, is also plotted
in Figure 4.51. Using the method introduced in Equation 4.24, the average
degree of consolidation is estimated to be 42% (Chu et al., 2009). If we
use settlement data and apply Asaoka’s method to Figure 4.48b, the average
degree of consolidation is calculated as 91% (Chu et al., 2009), which was
overestimated.
Field vane shear and CPT tests were conducted 14 months after the appli-
cation of the surcharge. A comparison of the undrained shear strength pro-
iles obtained from ield vane shear tests conducted before and 14 months
after surcharge as well as from CPT (with pore pressure measurement)
tests conducted 14 months after surcharge is shown in Figure 4.52. Note
that the ground had settled for more than 2 m within 14 months, as shown
in Figure 4.50a. This explains why the starting points of the in-situ tests
conducted before and after the surcharge are different. In Figure 4.52, the
undrained shear strength proile estimated by assuming a uniform degree
of consolidation of 90% is also plotted for comparison. As mentioned, a
silty sand was present at an elevation of −12.5 mCD and a sand blanket
 Prefabricated vertical drains 151

7
2nd PVD

Surcharge elevation (mCD)

installation

3
1st PVD
installation Surcharge elevation
Water level
1
0 90 180 270 360 450 540
(a) Time (days)

0.5
Settlement (m)

1.5

2.5 SP-506 DS-503

DS-504 DS-505
DS-511 DS-507
3
0 90 180 270 360 450 540
(b) Time (days)

150
Excess pore pressure (kPa)

125

100

75
PZ-47
50 PZ-48
PZ-49
25

0
0 90 180 270 360 450 540
(c) Time (days)

Figure 4.50 Monitoring data during the reclamation of slurry pond. (a) Surcharge varia-
tion versus time. (b) Settlement versus time. (c) Excess pore pressure ver-
sus time. Initial positions of the instruments: SP-506 at +3.5 mCD; DS-503
at −5.3 mCD; PZ-47, DS-504 and PP-466 at −8 mCD; PZ-48, DS-505 and
PP-467 at −10 mCD; PZ-49, DS-511 and PP-468 at −12 mCD. (From Chu, J.,
Bo, M.W. and Choa, V. (2006). Geotextiles and Geomembranes, Vol. 24, No. 6,
pp. 339–348.)
152 Ground improvement

Pore water pressure (kPa)

0 100 200 300
–5

–6

–7

–8
Elevation (m)

–9

–10

–11
Hydrostatic
–12 Surcharge
At 16 mths

–13

Figure 4.51 Pore water pressure distribution based on Figure 4.50c.

0 10 20 30 40 50 60 70 80 90 100
–7

–8 FVT (prior to
reclamation)

–9 FVT
Elevation (mCD)

CPT
–10
90% DOC

–11

–12

–13
Shear strength (kPa)

Figure 4.52 Comparison of undrained shear strength proiles measured before and after
consolidation for 14 months. (Redrawn from Chu, J., Bo, M.W. and Choa, V.
(2006). Geotextiles and Geomembranes, Vol. 24, No. 6, pp. 339–348.)
 Prefabricated vertical drains 153

was placed on the top surface. Therefore, there was combined vertical and
horizontal drainage near the top and bottom boundaries. This explains
why the shear strength increment was the largest at both the top and the
bottom. The consolidation in the middle of the clay layer ranging from
−9.5 to −12.0 m was contributed mainly by horizontal drainage to the
PDVs. Applying a method similar to Equation 4.24, the degree of consoli-
dation can be estimated as 45%, which is similar to that based on pore
water pressure.

4.7.2 Combined fill and vacuum preloading case

The second case is for the soil treatment of a storage yard in Tianjin Port,
China. The storage yard was located on a 16-m-thick soft clay layer. The
top 3–4 m of the clay layer was reclaimed recently using clay slurry dredged
from seabed. The remaining 16–19 m was original seabed clay. The soil
in both layers was soft and still undergoing consolidation. This soft clay
layer needed to be improved before the site could be used as a storage yard.
Preloading using ill surcharge alone was not feasible as it was dificult to
place a ill embankment several meters high on soft clay. The vacuum pre-
loading method could be used. However, the nominal vacuum load of 80
kPa was not suficient for this project. Therefore, a combined vacuum and
ill surcharge preloading method was adopted. Fill surcharge of a height
ranging from 2.53–3.50 m was applied in addition to the vacuum load and
0.3 m of sand blanket. The ill was applied in stages partially for stability
consideration and partially due to practical constraints in transporting ill.
The ill used was a silty clay with an average unit weight of 17.1 kN/m3.
The layout of the storage yard is shown in Figure 4.53. It was an L-shape
with a total area of 7433 m 2 . For the convenience of construction, the site
was divided into three sections, I, II and III, as shown in Figure 4.53. The
water content of the soil was higher than or as high as the liquid limit at
most locations in the soft clays. The ield vane shear strength of the soil was
generally 20– 40 kPa. For more information on the site conditions and soil
properties, see Yan and Chu (2007).
The soil improvement work was carried out as follows. A 0.3-m sand
blanket was irst placed on the ground surface. PVDs were then installed
on a square grid at a spacing of 1.0 m to a depth of 20 m. Corrugated lex-
ible pipes (100 mm diameter) were laid horizontally in the sand blanket to
link the PVDs to the main vacuum pressure line. The pipes were perforated
and wrapped with a permeable fabric textile to act as a ilter layer. Three
layers of thin PVC membrane were laid to seal each section. Vacuum pres-
sure was then applied using jet pumps. The schematic arrangement of the
vacuum preloading method used is similar to that shown in Figure 4.44.
The soil improvement started from Section I, followed by Section II and
then Section III. The loading sequence and the ground settlements induced
154 Ground improvement

W E
I II III

30.0 m

50.0 m
80.0 m 119.0 m 27.881 m

Settlement gauge Field vane

PWP transducer Inclinometer

Multi-level gauge Piezometer
Borehole

Figure 4.53 Project site layout and plan view of instrumentation. (Redrawn from Yan,
S.W. and Chu, J. (2005). Canadian Geotechnical Journal, Vol. 42, No. 4,
pp. 1094–1104.)

by the vacuum and surcharge loads for Section II are shown in Figure 4.54.
The vacuum load was applied for 4–8 weeks before ill surcharge loads
were applied in stages. The total ill height applied was 3.5 m for Section
II. The maximum surface settlement induced by the vacuum and surcharge
loads in this section was 1.614 m.

160 Total pressure

120
Preload (kPa)

80
Vacuum pressure under membrane

40
Duration (day)
0
30 60 90 120 150 180
Settlement (cm)

–40
Zone II
–80 Surcharge h = 3.50 m

–120

–160

Figure 4.54 Loading sequence and ground settlement measured at Section II. (Redrawn
from Yan, S.W. and Chu, J. (2005). Canadian Geotechnical Journal, Vol. 42,
No. 4, pp. 1094–1104.)
 Prefabricated vertical drains 155

Instruments including surface settlement plates, multi-level settlement

gauges, and standpipes were installed in all three sections to monitor the
consolidation performance. However, inclinometers were installed in sec-
tions I and II only and pore water pressure transducers in Section II only
due to budget constraints. The locations of those instruments are shown
schematically in Figure 4.53 (plan view) and Figure 4.55 (elevation view).
Soil samples were taken from sections II and III both before and after soil
improvement for laboratory tests. Field vane shear tests were also con-
ducted at sections II and III both before and after the soil improvement.
Some settlements took place after the vertical drains were installed, but
before the vacuum and surcharge loads were applied. The durations between
the installation of vertical drains and the application of vacuum loads were
3–4 weeks. The ground settlement measured before the application of vac-
uum loads was 0.21, 0.31, and 0.25 m for sections I, II, and III respectively.
The settlements were induced mainly as a result of the dissipation of the
existing excess pore water pressures in the soil as the soil was still under
consolidation due to land reclamation. The disturbance to the soil caused
by the installation of the vertical drains also contributed to the settlement.
The settlements monitored by the settlement gauges installed at differ-
ent depths during vacuum and surcharge loadings are plotted versus dura-
tion for Section II in Figure 4.56. Settlements were observed in soil up to
14.5 m deep, which indicates that the vacuum preloading was effective for

Multi-level gauge PWP transducer Inclinometer Piezometer

0.1

3.0
4.5

10.0
6.0
7.0

9.0
10.5

12.0
12.5

14.5 15.0

16.5 16.5

Figure 4.55 Elevation view of instrumentation. (Redrawn from Yan, S.W. and Chu, J.
(2005). Canadian Geotechnical Journal, Vol. 42, No. 4, pp. 1094–1104.)
156 Ground improvement

Duration (days)
0 20 40 60 80 100 120 140 160 180 200
0.0

–0.2
Settlement (m)

–0.4

–0.6
3.8
7.0
–0.8 10.5
12.5
14.5
–1.0

Figure 4.56 Settlement measured at different depths against duration at Section II. (The
arrows represent where staged loads were applied)

Duration (days)
0 20 40 60 80 100 120 140 160 180 200
0.0
+1.0
–2.0
–20.0 –5.0
–8.0
Pore water pressure

–11.0
reduction (kPa)

–40.0 –12.5

–60.0

–80.0

–100.0

Figure 4.57 Pore water pressure reductions measured at different depths against dura-
tion at Section II (the arrows represent where staged loads were applied).
(Redrawn from Yan, S.W. and Chu, J. (2005). Canadian Geotechnical Journal,
Vol. 42, No. 4, pp. 1094–1104.)
 Prefabricated vertical drains 157

the entire 16-m soft clay. The reductions in the pore water pressures mea-
sured by the piezometers installed at different depths are plotted versus
loading duration in Figure 4.57 for Section II. Under the vacuum load, the
pore water pressures reduced quickly with time. However, when the ill sur-
charge was applied, a localised pore pressure increase occurred. The times
at which staged surcharge loads were applied are indicated by arrows in
Figure 4.57. It can be seen that the localised increase in pore water pressure
coincides with the application of surcharge loads.
Based on the pore water pressure monitoring data shown in Figure
4.57, the pore water pressure distributions with depth at the initial stage,
30 and 60 days, and the inal stage are plotted in Figure 4.58. Before
the application of vacuum and surcharge loads, the initial pore water

Pore water pressure (kPa)

–100 0 100 200 300
0
Initial

2 Static
Suction

4 Final
30 days

6 60 days
u0 + ∆σ

8
Depth (m)

u0 (z) + ∆σ

10
u0 (z)

12
us

14

16

uf (z)
18

Figure 4.58 Pore water pressure distributions with depth at Section II. (Redrawn from
Yan, S.W. and Chu, J. (2005). Canadian Geotechnical Journal, Vol. 42, No. 4,
pp. 1094–1104.)
158 Ground improvement

Table 4.9 Ultimate settlement and degree of consolidation

estimated by different methods
Asaoka’s method Based on pore pressure
Section S∞ (m) Uf (%) Uf (%)
II 1.84 87 82

pressures, u 0(z), were greater than the hydrostatic pore water pressure.
The total ill surcharge was about 60 kPa for Section II. Using the verti-
cal stress calculated, the initial pore water pressure distribution after the
application of the ill surcharge is shown as u 0(z) + Δσ in Figure 4.58.
The suction line for a suction of −80 kPa is also plotted in Figure 4.58 as
the line u s. The pore water pressure distributions at 30 and 60 days and
the end of preloading (uf (z)) are also shown in Figure 4.58. These curves
show the changes of the pore water pressure proiles with time. The area
bound by the inal pore water pressure curve, uf (z), and the suction line,
u s , represents the remaining excess pore water pressures that have not
dissipated.
The degree of consolidation can be estimated using either settlement or
pore water pressure data. For the former, Asaoka’s method was applied
to predict the ultimate settlements, S ∞, using the ground settlement data
shown in Figure 4.54. The results are given in Table 4.9. Using the pore
water pressure distribution proile shown in Figure 4.58 and Equation
4.24, the average degree of consolidation at the end of preloading, Uf, can
be estimated as 82%. The reasons why the degree of consolidation esti-
mated based on settlement is higher than that based on pore water pressure
were explained in Yan and Chu (2005).
Field vane shear tests were conducted before and after preloading in
Section II and the results are presented in Figure 4.59. It can be seen that
considerable improvement in the vane shear strength was achieved through-
out the entire depth of 16 m where ield vane shear tests were conducted.
On average, the vane shear strength increased twofold.

4.8 SUMMARY

A comprehensive review of the recent development in theories and practice

related to the use of PVDs for soil improvement is presented. These include
the outline of the fundamental consolidation theories, analytical methods
for the design of PVDs, the determination of soil parameters for both intact
and smeared soil, the types of PVDs, and the selections of PVDs. Issues
related to construction such as installation equipment, quality-control
tests, measurement of penetration depth, ield instrumentation, and meth-
ods for the performance evaluation of PVDs, design codes, standards, and
 Prefabricated vertical drains 159

Vane shear strength (kPa)

0 20 40 60 80 100 120
0

Section II

8
Depth (m)

12

16

Before
After

20
Figure 4.59 Field vane strength proile before and after soil improvement at Section II.
(Redrawn from Yan, S.W. and Chu, J. (2005). Canadian Geotechnical Journal,
Vol. 42, No. 4, pp. 1094–1104.)

speciications are also discussed. Some speciic methods for using PVDs for
vacuum preloading are also introduced. Finally, two case studies, one for
the use of PVDs for soil improvement of slurry type of soil and another for
the use of PVDs for a combined ill and vacuum surcharge for the improve-
ment of the soft soil below a storage yard, are presented to illustrate the
practical applications of PVDs for different soil improvement projects. In
conclusion, suficient research development has been made and practical
experience been gained over the years on the use of PVDs for soil improve-
ment. However, it is still more of an art than science as far as for the predic-
tion of the outcome of a soil improvement scheme is concerned. A holistic
approach is thus advocated for the implementation of PVD techniques for
soil improvement. The design methods proposed should be used together
160 Ground improvement

with careful ield observations and interpretation of soil instruments and

data obtained from in-situ and laboratory tests. Attention should be paid
to local knowledge of the soil conditions and implementation techniques.
A sensible interpretation of the acceptance criteria of the soil improvement
and the quality assurance of PVDs are among some of the important con-
siderations. As in all civil engineering projects, a good working relationship
between the supervising engineers and the contractors who are responsible
for the installation of the drains is essential to ensure the eventual success
of the project.

ACKNOWLEDGMENTS

Some of the work presented in this chapter formed part of a research pro-
gramme at the Nanyang Technological University, Singapore. The contri-
butions of Prof. Victor Choa, Dr. Myint Win Bo, Prof. Bak Kong Low,
Prof. A. Arulrajah, Dr. Daping Xiao and Haojie Liu to the programme are
gratefully acknowledged. The irst author has beneited greatly from col-
laborations with Prof. Shuwang Yan, Prof. Hanlong Liu, Prof. Buddhima
Indraratna, Prof. Dennes Bergado, Prof. Pedro Pinto, Prof. Robert Lo,
Prof. Cholachat Rujikiatkamjorn, Kok Pang Lam, Serge Varaksin and
Kenny Yee. Their support and cooperation is also much appreciated.
Finally the writers would like to thank Dr. Zhiwei He for proofreading
the manuscripts.

REFERENCES

Akagi, T. (1994). Hydraulic applications of geosynthetics to iltration and drain-

age problems – with special reference to prefabricated band-shaped drains,
Proc. 5th Int. Conf. on Geotextiles, Geomembranes and Related Products,
Singapore, preprint of keynote lecture, pp. 99–119.
Arulrajah, A., Bo, M.W. and Chu, J. (2009). Instrumentation at the Changi land
reclamation project, Singapore. Geotechnical Engineering, Proceedings of the
Institution of Civil Engineers, Vol. 162, No. 1, pp. 33–40.
Arulrajah, A., Nikraz, H., and Bo, M.W. (2004). Factors affecting ield instrumen-
tation assessment of marine clay treated with prefabricated vertical drains,
Geotextiles and Geomembranes, Vol. 22, No. 5, pp. 415–437.
Asaoka, A. (1978). Observational procedure of settlement prediction. Soils and
Foundations, Vol. 18, No. 4, pp. 87–101.
ASTM D4491-99a. (2009). Standard Test Methods for Water Permeability of
Geotextiles by Permittivity. ASTM Standards.
ASTM D4632. (2008). Standard Test Method for Grab Breaking Load and
Elongation of Geotextile. ASTM Standards.
 Prefabricated vertical drains 161

ASTM D4716. (2008). Standard Test Method for Determining the (In-plane) Flow
Rate per unit Width and Hydraulic Transmissivity of a Geosynthetic Using a
Constant Head, ASTM Standards.
ASTM D4751. (2004). Standard Test Method for Determining Apparent Opening
Size of a Geotextile, ASTM Standards.
AS8700 (2011) Execution of Prefabricated Vertical Drains, Standards Australia.
Balasubramaniam, A.S., Bergado, D.I., Long, P.V. and Thayalan. (1995). Experiences
with sand drains and prefabricated vertical drains in ground improvement of
soft clays, Seminar on Engineering for Coastal Development, Singapore.
Barron, R.A. (1948). Consolidation of ine-grained soils by drain wells, Trans ASCE,
Vol. 113, Paper 2346, pp. 718–54.
Bergado, D.T., Singh, N., Sim, S.H., Panichayatum, B., Sampaco, C.L. and
Balasubramaniam, A.S. (1990). Improvement of soft Bangkok clay using ver-
tical geotextile band drains compared with granular piles, Geotextiles and
Geomembranes, Vol. 9, No. 3, pp. 203–231.
Bergado, D.T., Sampaco, C.L., Shivashankar, R., Alfaro, M.C., Anderson, L.R. and
Balasubramaniam, A.S. (1991). Performance of a welded wire wall with poor
quality backills on soft clay, Proc. ASCE Geotechnical Congress, Boulder,
Colorado, pp. 909–922.
Bergado, D.T., Alfaro, M.C., and Balasubramaniam, A.S. (1993a). Improvement of
soft Bangkok clay using vertical drains, Geotextiles and Geomembranes, Vol.
12, No. 7, pp. 615–663.
Bergado, D.T., Mukherjee, K., Alfaro, M.C. and Balasubramaniam, A.S. (1993b).
Prediction of vertical-band-drain performance by the inite-element method,
Geotextiles and Geomembranes, Vol. 12, No. 6, pp. 567–586.
Bergado, D.T., Long, P.V., Lee, C.H., Loke, K.H. and Werner, G. (1994). Performance
of reinforcement embankment on soft Bangkok clay with high strength geo-
textile reinforcement, Geotextiles and Geomembranes, Vol. 13, pp. 403–420.
Bergado, D.T., Manivannan, R., Balasubramaniam, A.S. (1996a). Proposed cri-
teria for discharge capacity of prefabricated vertical drains, Geotextiles and
Geomembranes, Vol. 14, pp. 481–505.
Bergado, D.T., Anderson, L.R., Miura, N. and Balasubramaniam, A.S. (1996b). Soft
Ground Improvement in Lowland and Other Environments, ASCE Press,
ASCE, New York.
Bergado, D.T., Balasubramaniam, A.S., Fannin, R.J. and Holtz, R.D. (2002).
Prefabricated vertical drains (PVDs) in soft Bangkok clay: a case study of the
New Bangkok International Airport project, Canadian Geotechnical Journal,
Vol. 39, pp. 304–315.
Bo M.W., Chu, J. and Choa, V. (1999). Factors affecting the assessment of degree
of consolidation, Proc. International Conference on Field Measurements in
Geomechanics, Balkema, Rotterdam, pp. 481–486.
Bo, M.W., Chu, J., Low, B.K. and Choa, V. (2003). Soil Improvement: Prefabricated
Vertical Drain Technique, Thomson Learning, Singapore.
Bo, M.W. (2004). Discharge capacity of prefabricated vertical drain and their ield
measurements, Geotextiles and Geomembranes, Vol. 22, No. 1–2, pp. 37–48.
Bo, M.W., Chu, J., Choa, V. (2005). Changi East Reclamation and Soil Improvement
Project. In: Chapter 9, Ground Improvement—Case Histories. Indraratna, B.,
Chu, J. (Eds.), Elsevier, Amsterdam, pp. 247–276.
162 Ground improvement

Broms, B.B., Chu, J., and Choa, V. (1994). Measuring the discharge capacity of band
drains by a new drain tester, Proc. 5th Int. Conf. on Geotextiles, Geomembranes
and Related Products, Vol. 3, Singapore, pp. 803–806.
BS EN 15237. (2007). European Standard on Execution of Special Geotechnical
Works—Vertical Drainage. European Standard.
Carillo, N. (1942). Simple two- and three-dimensional cases in the theory of consoli-
dation of soils. Journal of Mathematics and Physics, Vol. 21, No. 1, 1e5.
Carroll, R.G., (1983). Geotextile ilter criteria. Transportation Research Record, No.
916, pp. 46–53.
Chai, J.C., Miura, N., Sakajo, S. and Bergado, D.T. (1995). Behaviour of vertical
drain improved subsoil under embankment loading, Soils and Foundations,
Vol. 35, No. 4, pp. 49–61.
Chai, J.C., Shen, S.L., Miura, N. and Bergado, D.T. (2001). Simple method of model-
ling PVD improved subsoil, Journal of Geotechnical and Geoenvironmental
Engineering, ASCE, Vol. 127, No.11, pp. 965–972.
Choa, V., Bo, M.W., and Chu, J., (2001). Soil improvement works for Changi East
reclamation project. Ground Improvement, Vol. 5, No. 4, pp. 141–153.
Choa, V., Vijiaratnam, A. Karunaratne, G.P., Ramaswamy, S.D. and Lee, S.L. (1979).
Pilot test for soil stabilisation at Changi Airport, Proceedings 7th European
Regional Conference on Soil Mechanics and Foundation Engineering, Brighton,
Vol. 3, pp. 29–36.
Chu, J. and Choa, V. (1995). Quality control tests of vertical drains for a land
reclamation project, Compression and Consolidation of Clayey Soils,
In: Yoshikuni, H., Kusakabe, O. (Eds.), Proceedings of an International
Symposium, Hiroshima, Japan, 10–12 May, A.A. Balkema, Rotterdam, pp.
43–48.
Chu, J., Yan, S.W. and Yang, H. (2000). Soil improvement by the vacuum pre-
loading method for an oil storage station. Géotechnique, Vol. 50, No. 6, pp.
625–632.
Chu, J., Bo, M.W., Chang, M.F., and Choa, V. (2002). Consolidation and per-
meability properties of Singapore marine clay, Journal Geotechnical and
Geoenvironmental Engineering, ASCE, Vol. 128, No. 9, pp. 724–732.
Chu, J., Bo, M.W. and Choa, V. (2004). Practical consideration for using vertical
drains in soil improvement projects, Geotextiles and Geomembranes, Vol. 22,
pp. 101–117.
Chu, J. and Yan, S.W. (2005a). Estimation of degree of consolidation for vacuum
preloading projects, International Journal of Geomechanics, ASCE, Vol. 5, No.
2, pp. 158–165.
Chu, J. and Yan, S.W. (2005b) Application of the vacuum preloading method in
land reclamation and soil improvement projects. Chapter 3, In Ground
Improvement—Case Histories, B. Indraratna and J. Chu, (Eds.), Elsevier, pp.
91–118.
Chu, J., Goi, M.H. and Lim, T.T. (2005). Consolidation of cement treated sewage
sludge using vertical drains. Canadian Geotechnical Journal, Vol. 42, No. 2,
pp. 528–540.
Chu, J., Bo, M.W. and Choa, V. (2006). Improvement of ultra-soft soil using pre-
fabricated vertical drains, Geotextiles and Geomembranes, Vol. 24, No. 6,
pp. 339–348.
 Prefabricated vertical drains 163

Chu, J., Varaksin, S., Klotz, U. and Mengé, P. (2009a). Construction processes.
State-of-the-Art-Report, Proc. 17th International Conf on Soil Mechanics and
Geotechnical Engineering, Alexandria, Egypt, 5–9 Oct., Vol. 4, pp. 3006–3135.
Chu, J., Bo, M.W., and Arulrajah, A. (2009b). Reclamation of a slurry pond in
Singapore, Geotechnical Engineering, Proceedings of the Institution of Civil
Engineers, Vol. 162, No. GE1, pp. 13–20.
Chu, J., Bo, M.W. and Arulrajah, A. (2009c). Soil improvement works for an off-
shore land reclamation. Proceedings of the Institution of Civil Engineers—
Geotechnical Engineering, Vol. 162, No. 1, pp. 21–32.
Chu, J., Indraratana, B., Yan, S. and Rukikiatkamjorn, C. (2012). Soft soil improve-
ment through consolidation: an overview. State-of-the-art Report, Proc.
International Conference on Ground Improvement and Ground Control,
Vol. 1 (in print).
Collazos, O.M., Bowders, J.J. and Bouazza, A. (2002). Prefabricated vertical drains
for use in soil vapor extraction applications. Transportation Research Record,
No. 1786, pp. 104–111.
Duncan, J.M. (1993). Limitations of conventional analysis of consolidation settle-
ment, 27th Karl Terzaghi Lecture, Journal Geotechnical and Geoenvironmental
Engineering, ASCE, Vol. 119, No. 9, pp. 1333–1359.
Fang, Z. and Yin, J.H. (2006). Physical modelling of consolidation of Hong Kong
marine clay with prefabricated vertical drains. Canadian Geotechnical Journal,
Vol. 43, pp. 638–652.
Goi, M.H. (2004). Use of stabilised sewage sludge for land reclamation, MEng the-
sis, Nanyang Technological University, Singapore.
Guido, V. A. and Ludewig, N. M. (1986). ‘A compative laborotry evaluation of band-
shaped prefabricated drains’. Consolidation of Soils: Testing and Evaluation.
ASTM STP 892, Yong, R. N. and Townsend, F. C. Eds. 642–662.
Hansbo, S. (1960). Consolidation of clay, with special reference to inluence of verti-
cal sand drains, Proc. Swedish Geotechnical Institute 18, Linkoping.
Hansbo, S. (1979). Consolidation of clay by band-shaped prefabricated drains,
Ground Engineering, Vol. 12, No. 5, pp. 16–25.
Hansbo, S. (1981). Consolidation of ine-grained soils by prefabricated drains, Proc.
10th International Conference Soil Mechanics and Foundation Engineering,
Stockholm, Vol. 3, pp. 677–682.
Hansbo, S. (1983). How to evaluate the properties of prefabricated drains, Proc. 8th
European Conf. on Soil Mechanics and Foundation Engineering, Helsinki, Vol.
2, pp. 621–626.
Hansbo, S. (1997). Aspects of vertical drain design: Darcian or non-Darcian low,
Géotechnique, Vol. 47, pp. 983–992.
Hansbo, S. (2001). Consolidation equation valid for both Darcian or non-Darcian
low, Géotechnique, No. 1, pp. 51–54.
Hansbo, S. (2004). Band drains. In Ground Improvement (2nd edition), Spon Press,
pp. 4–56.
Hansbo, S. (2005). Experience of consolidation process from test areas with and
without vertical drains. In B. Indraratna and J. Chu (Eds.), Chapter 1, Ground
Improvement—Case Histories, Elsevier, Oxford, pp. 3–50.
Hight, D.W., Bond, A.J. and Legge, J.D. (1992). Characterisation of Bothkannar clay:
an overview, Géotechnique, Vol. 42, No. 2, pp. 303–347.
164 Ground improvement

Hird, C.C., Pyrah, I.C., and Russell, D. (1992). Finite element modelling of vertical
drains beneath embankments on soft ground, Géotechnique, Vol. 42, No. 3,
pp. 499–511.
Hird, C.C. and Moseley, V.J. (2000). Model study of seepage in smear zones around
vertical drains in layered soil, Géotechnique, Vol. 50, No. 1, pp. 89–97.
Holtz, R.D. and Holm, B.G. (1972). Excavation and sampling around some sand
drains at Ska-Edeby, Sweden, Proc. 6th Scandinavian Geotechnical Meeting,
Trondheim, Norwegian Geotechnical Institute, pp. 75–89.
Holtz, R.D. and Kovacs, W.D. (1981). An Introduction to Geotechnical Engineering,
Prentice-Hall, Englewood Cliffs, New Jersey.
Holtz, R.D. (1987). Preloading with prefabricated vertical strip drains, Geotextiles
and Geomembranes, Vol. 6, No. 1–3, pp. 109–131.
Holtz, R.D., Jamiolkowski, M., Lancellotta, R. and Pedroni, R. (1991). Prefabricated
Vertical Drains: Design & Performance, CIRIA Ground Engineering Report,
Butterworth-Heinemann Ltd., London.
Holtz, R.D., Shang, J.Q. and Bergado, D.T. (2001). Soil improvement, in Geotechnical
and Geoenvironmental Engineering Handbook, R. K. Rowe (Ed.), pp. 429–462.
ICE. (1982). Vertical Drains, Thomas Telford Ltd., London.
Indraratna, B., Balasubramaniam, A.S. and Ratnayake, P. (1994). Performance of
embankment stabilized with vertical drains on soft clay, Journal of Geotechnical
Engineering, ASCE, Vol. 120, No. 2, pp. 257–273.
Indraratna, B. and Redana, I.W. (1997). Plane strain modelling of smear effects asso-
ciated with vertical drains, Journal of Geotechnical and Geoenvironmental
Engineering, ASCE, Vol. 123, No. 5, pp. 474–478.
Indraratna, B. and Redana, I.W. (1998). Laboratory determination of smear zone
due to vertical drain installation, Journal Geotechnical Engineering, ASCE,
Vol. 124, No. 2, pp. 180–184.
Indraratna, B. and Redana, I.W. (2000). Numerical modelling of vertical drains with
smear and well resistance installed in soft clay, Canadian Geotechnical Journal,
Vol. 37, pp. 132–145.
Indraratna, B., Chu, J. (Eds.) (2005). Ground Improvement Case Histories, Elsevier,
Oxford, 1115 pp.
Indraratna, B., Rujikiatkamjorn, C., Balasubramaniam, A.S., and Wijeyakulasuriya,
V. (2005). Predictions and observations of soft clay foundations stabilised with
geosynthetic drains and vacuum surcharge. In: Indraratna, B. and Chu, J. (Eds.),
Ground Improvement—Case Histories. Elsevier, Amsterdam, pp. 199–229
(Chapter 7).
Jamiolkowski, M., Lancellotta, R. and Wolski, W. (1983). Precompression and
speeding up consolidation, General Report—Specialty Session 6, Proc. 8th
European Soil Mechanics and Foundation Engineering, Helsinki, Vol. 3, pp.
1201–1226.
JTJ/T256-96. (1996). Code of Practice for Installation of Prefabricated Drains.
Ministry of Communications, China.
JTJ/T257-96. (1996). Quality Inspection Standard for Prefabricated Drains. Ministry
of Communications, China.
Kamon, M., Pradhan, B.S., and Suwa, S. (1984). Laboratory evaluation of the pre-
fabricated band-shaped drains, soil improvement, Current Japanese Materials
Research, Vol. 9, Cambridge University Press, Cambridge, UK.
 Prefabricated vertical drains 165

Karunaratne, G.P. (2011). Prefabricated and electrical vertical drains for con-
solidation of soft clay, Geotextiles and Geomembranes, Vol. 29, No. 4, pp.
391–401.
Kitazume, M. (2007). Design, execution and quality control of ground improve-
ment in land reclamation, Proc. of the 13th Asian Regional Conference on Soil
Mechanics and Foundation Engineering, keynote lecture.
Kjellman, W. (1948). Accelerating consolidation of ine grained soils by means of
cardboard wicks. Proc. 2nd International Conference on Soil Mechanics and
Foundation Engineering. Vol. 2, Rotterdam, pp. 302–305.
Kolff, A.H.N., Spierenburg, S.E.J. and Mathijssen, F.A.J.M. (2004). BeauDrain: a
new consolidation system based on the old concept of vacuum consolidation,
Proc. 5th International Conference on Ground Improvement Techniques,
Kuala Lumpur, Malaysia.
Kremer, R.R.H.J., Oostven, J.P., Van Weele, A.F., Dejager, W.F.J., and Meyvogel, I.J.
(1983). The quality of vertical drainage, Proc. 8th European Conf. on Soil
Mechanics and Foundation Engineering, Vol. 2, Helsinki, pp. 721–726.
Lee, S.L., Karunaratne, G.P. and Aziz, M.A. (2003). Design and performance of
Fibredrain in soil improvement projects. Ground Engineering, Vol. 10, No. 4,
pp. 149–156.
Lee, C.H. and Kang, S.T. (1996). Discharge capacity of prefabricated vertical band
drains, Final Year Report, Nanyang Tech. Univ., Singapore.
Leroueil, S. (1996). Compressibility of clays: fundamental and practical aspects.
Journal Geotech. Engrg. Div., ASCE, Vol. 122, No. 7, pp. 534–543.
Li, A.L. and Rowe, R.K. (2001). Combined effects of reinforcement and prefab-
ricated vertical drains on embankment performance, Canadian Geotechnical
Journal, Vol. 38, No. 6, pp. 1266–1282.
Liu, H.L. and Chu, J. (2009). A new type of prefabricated vertical drain with improved
properties, Geotextiles and Geomembranes, Vol. 27, No. 2, pp. 152–155.
Liu, H.L., Chu, J. and Ren, Z.Y. (2009). New methods for measuring the installation
depth of prefabricated vertical, Geotextiles and Geomembranes, Vol. 27, No.
6, pp. 493–496.
Matyas, E.L. and Rothenburg, L. (1996). Estimation of total settlement of embank-
ments by ield measurements, Canadian Geotechnical Journal, Vol. 33, pp.
834–841.
Mesri, G and Castro, A. (1987). The Cα/Cc concept and K0 during secondary com-
pression. Geotech. Engrg. Div., ASCE, Vol. 112, No. 3, pp. 230–247.
Mesri, G. and Lo, D.O.K. (1991). Field performance of prefabricated vertical drains,
Proc. Int. Conf. on Geotechnical Engineering for Coastal Development,
Yokohama, Vol. 1, pp. 231–236.
Mesri, G., Kjlouni, M.A., Feng, T.W. and Lo, D.O.K. (2001). Surcharging of soft
ground to reduce secondary settlement. Soft Soil Engineering, Lee, C.F. et al.
(Eds.), pp. 55–65.
Mitchell, J.K. and Katti R.K. (1981). Soil improvement—state of the art report,
Proceedings 10th International Conference Soil Mechanics and Foundation
Engineering, Stockholm, Vol. 4, pp. 509–565.
Moseley, M.P. and Kirsch, K. (2004). Ground Improvement (2nd ed.), Spon Press.
Nakanado, H., Hanai, H. and Imai, W. (1992). Permeability of plastic board drains.
Technical Report, Vol. 17, Hukken, Co. Ltd., pp. 139–151.
166 Ground improvement

Onoue, A., Ting, N.H., Germaine, J.T. and Whitman, R.V. (1991). Permeability of
disturbed zone around vertical drains, Proc. of ASCE Geotechnical Engineering
Congress, Colorado, pp. 879–890.
Park, Y.M. and Miura, N. (1998). Soft ground improvement using prefabricated
vertical drains (PVD). In: Miura, N., Bergado, D. (Eds.), Improvement of Soft
Ground, Design, Analysis, and Current Researches, pp. 35–48.
Park, J.H., Yuu, J. and Jeon, H.Y. (2010). Green geosynthetics applications to sus-
tainable environmental ields from the viewpoint of degradability, Proceedings
International Symposium and Exhibition on Geotechnical and Geosynthetics
Engineering: Challenges and Opportunities on Climate Change, 7–8 Dec
Bangkok, pp. 43–50.
Raison, C. (2004). Ground and Soil Improvement: Thomas Telford, London.
Ren, Z.Y. (2004). A prefabricated vertical drain with capability for penetration depth
measurement, China patent No. 0119469.0.
Rujikiatkamjorn, C., Indraratna, B. and Chu, J. (2008). 2D and 3D Numerical
Modeling of Combined Surcharge and Vacuum Preloading with Vertical
Drains, International Journal of Geomechanics, Vol. 8, No. 2, 144–156.
Saowapakbiboon, J., Bergado, D.T., Chai, J.C., Kovittayanon, N., and de Zwart,
T.P. (2008). Vacuum-PVD combination with embankment loading consolida-
tion in soft Bangkok clay: A case study of Suvarnabhumi Airport project,
Proc. 4th Asian Regional Geosynthetics Conference, Shanghai, China, pp.
440–449.
Schaefer, V.R., Abramson, L.W., Drumheller, J.C., Hussion, J.D., & Sharp, K.D.
(1997). Ground improvement, Ground Reinforcement and Ground Treatment:
Developments 1987–1997, ASCE GSP 69.
Seah, T.H. (2006). Design and construction of ground improvement works at
Suvarnabhumi Airport. Geotechnical Engineering Journal of Southeast Asian
Geotechnical Society, Vol. 37, pp. 171–188.
Shang, J.Q. (1998). Electroosmosis-enhanced preloading consolidation via vertical
drains. Canadian Geotechnical Journal, Vol. 35, pp. 491–499.
Sivaram, B. and Swamee, P. (1977). A computational method for consolidation coef-
icient. Soils and Foundations, Vol. 17, No. 2, pp. 48–52.
Sridharan, A. and Sreepada Rao, A. (1981). Rectangular hyperbola itting method
for one-dimensional consolidation, Geotechnical Testing Journal, Vol. 4, No.
4, pp. 161–168.
Suits, L.D., Gemme, R.L. and Masi, J.J. (1986). Effectiveness of prefabricated drains
on laboratory consolidation of remolded soils. In: Yong, R.N., Townsend, F.C.
(Eds.), Consolidation of Soils: Testing and Evaluation, ASTM STP 892, pp.
663–683.
Towhata, I. (2008). Geotechnical Earthquake Engineering, Springer.
Varaksin, S. & Yee, K. (2007). Challenges in ground improvement techniques for
extreme conditions: Concept and Performance, Proc. 16th Southeast Asian
Geotechnical Conference, 8–11 May: Kuala Lumpur, pp. 101–115.
Wang, T-R. and Chen, W-H. (1996). Development in application of prefabricated
drains in treatment of soft soils, General Report, Proc. 3rd Symp. On Weak
ground Improvement using PVDs, Oct., Lianyuangang, China, pp. 13–40.
Xiao, D.P. (2002). Consolidation of soft clay using vertical drains, PhD thesis,
Nanyang Tech. Univ., Singapore.
 Prefabricated vertical drains 167

Xie, K.H. (1987). Consolidation theories and optimisation design for vertical drains,
PhD Thesis, Zhejiang University, China.
Yan, S.W. and Chu, J. (2003). Soil improvement for a road using the vacuum pre-
loading method, Ground Improvement, Vol. 7, No. 4, pp. 165–172.
Yan, S.W. and Chu, J. (2005). Soil improvement for a storage year using the com-
bined vacuum and ill preloading method. Canadian Geotechnical Journal, Vol.
42, No. 4, pp. 1094–1104.
Yan, S.W., Chu, J., Fan, Q.J., and Yan, Y. (2009). Building a breakwater with prefab-
ricated caissons on soft clay. Proceedings of ICE, Geotechnical Engineering,
Vol. 162, No. 1, pp. 3–12.
Yin, J.-H. and Graham, J. (1999). Elastic visco-plastic modeling of the time-depen-
dent stress-strain behavior of soils. Canadian Geotechnical Journal, Vol. 36,
No. 4, pp. 736–745.
Chapter 5

Permeation grouting
Gert Stadler and Harald Krenn

CONTENTS

5.1 Introduction ................................................................................. 169

5.2 Preparatory works and design ...................................................... 175
5.2.1 Aspects of rheological laws for particulate grout mixes .... 176
5.3 Execution of works....................................................................... 188
5.4 Monitoring, controls and acceptance tests.................................... 193
5.5 Resources and equipment ............................................................. 194
5.6 Grout material .............................................................................. 195
5.7 Characteristics and applicability of chemical grouts..................... 197
5.7.1 Silicate grouts .................................................................... 198
5.7.2 Acrylamide grouts.............................................................. 200
5.7.3 Grouts having to fulil environmental standards ............... 200
5.8 Technical summary ...................................................................... 204
References ............................................................................................. 205

5.1 INTRODUCTION

Grouting techniques in general are intended to ill voids in the ground

(issures in rock and porosity in sediments) with the following aims:

r Increase resistance against deformation

r Supply cohesion, shear, and uniaxial compressive strength
r Reduce conductivity/transmissivity via interconnected porosities in
an aquifer (this is the most common goal)

Grouting uses luids like thin mortars, particulate suspensions, aque-

ous solutions, and chemical products like polyurethane, acrylates, or epoxy
injected into the ground under pressure, via boreholes and packers, providing
the geometrical layout of ‘points of attack’ in the underground space. By dis-
placing gas or groundwater, these luids ill pores and issures in the ground

169
170 Ground improvement

and thus—after setting and hardening within a predetermined lapse of time—

attribute new properties to the subsoil. The degree of saturation and the prop-
erties of the hardened grout do deine the degree of achieved improvement.
The irst applications of grouting were in the ields of mining (shaft-
sinking) and hydro-engineering (grouting under dams). These go as far
back as 1802, when Berigny repaired the foundation of a sluice at Dieppe
(France), followed by similar applications at Rochefort, where leaks into
a dock were stopped by mortar injections. Ground stabilisation around
city excavations for high-rise structures and subways (Metro) have been
prominently added to these examples, as well as immobilisation of waste,
grouting behind tunnel linings, and rehabilitation of concrete structures of
dams. A prominent example for the latter is shown in Figure 5.1.*
Commercial considerations and costs are, of course, at all times a matter
of the market, and therefore dificult to generalise. In general, grouting is
only viable if other more economical and ‘designable’ ground engineering
techniques would be physically impossible, and if the process may be accom-
plished within acceptable construction time using drilling techniques and
grout material both available and economical. Grouting pressures applied
must stay below the pressure level causing ground fracturing, and the tech-
nical result (for instance, increase in strength or reduction of permeability)
must be reasonably anticipated during design considerations.
Costs of grouting works are governed by technical and operative ‘boundaries’.
Typical performance rates are

r Average grouting rate per pump: 5–20 l/min

r Average man-hour [H] per operative pump-hour [h]: 1.1–3.5 H/h
r Average man-hour per ton of cement of a neat Ordinary Portland
Cement (OPC) grout: 5–10 H/ton
r Average minimum borehole spacing equal to the thickness of the
treatment, or <3 m
r Average percentage of voids on which to base the grout consumption:
in sediments, 22%–35%; in rocks, only 0.5%–1.5%
r Average metre of borehole per m3 of soil/rock grouted: 0.15–0.8 m/m3
r Average cost for depreciation (plus interest), and for maintenance
and repair of equipment and machinery: 3.6%–4.1% (of replacement
value) per month

* Structural repair of cracked concrete in a double curvature arch dam (the Koelnbrein Dam,
Figure 5.1) was accomplished by a specialized application of grouting with epoxy resins of high
viscosity and strength (applied under a considerable head of water!). Lombardi provided the
design for this repair work and took the occasion to apply his concept of a Grouting Intensity
Number (GIN) at this major repair project. Another typical example of grouting applications
is grouting of horizontal barriers (blankets) in sand below city excavations. To reduce seepage
during excavation of construction pits at gradients of as much as 10, it is possible to reduce per-
meabilities to around 1 × 10 -7 m/s, which corresponds to seepage values of 1.5 l/s per 1000 m2.
 Permeation grouting 171

KW-MALTA KOELNBREINSPERRE
MASSIGER GNEIS GEOLOGIE AN DER SPERRENSTELLE
ANSICHT VON DER LUFTSEITE
PLATTENGNEIS
SCHIEFERGNEIS
30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 09 08 07 06 05 04 03 02 01

1900

1800

1700

LAGE
y = –1000
18 17 16 15 14
13
19 12
1 20 11
222 10
23 9
x = –5000

24
N 26 R. 25 8
7
O CK 27 6
28 5
BL 29 4
30 0 50 100m 3
2
QUERSCHNITT 1

KGS

0 10 20m

Figure 5.1 Kölnbrein Dam, Austrian Drau Hydro Power Company, typical sections.

For a standard grouting application using these performance rates, suitable

estimates of total time and costs can be derived.
On top of these costs, approximately 25% of the cost of the operative person-
nel should be foreseen for supervision and infrastructural services. Mobilisation
and demobilisation costs are to be added to such approximate overall budget
considerations. These costs speciically depend on suitable access to the con-
struction site, conditions for transport, possibly conined working space, cli-
mate and time of the year, location of the drilling points, and other factors.
Measurement of and payment against grouting works should be based
on respective Bill of Quantities/Schedule of Rates, suficiently detailed to
address various operations and services related to grouting. For reference,
see the suggested standard schedule in Table 5.1, which could be adapted to
suit individual project situations.
Table 5.1 Proposed standard schedule of rates/ bill of quantities for grouting works

172
Item

Ground improvement
Item No Group Main Sub Unit Quantity Price price
1.01 Mobilization, site General site installation Ofices, store, personnel LS
installation, demobilization W/Shop, vehicles, etc
1.02 Rigs & equipment Mixing/batching plant No
1.03 Drill rig (Type) No
1.04 Grout pump (single, No
containerised)
1.05 Testing unit No
1.06 Additional units Type as above or other No
speciied
1.091 Relocating rigs & equipment Within project area Per item 1.02-1.06 No
(Type)
1.092 From site to site Per item 1.02-1.06 No
2.01 Rental of equipment General site installation Cal. week
2.02 Rigs & equipment Per item (as above) Cal. week
3.0 Idle/downtime (as speciied) Personnel (Category) on site Man hour
4.00 Setup & rigging drill over
hole
4.01 incl displacing rig > 2.5m No
4.02 Without displacement of rig No
5.00 Drilling for coring or grouting, in all types of ground, all directions and inclinations, collaring < 2.0m above working area
5.01 Coring 75-115mm
5.011 0-15m m
5.012 15-30m m
5.013 30-60m m
5.014 Extra over for casing m
5.02 Roto-Percussion drilling 56 bis 76 mm
5.021 0-6m m
5.022 6-12m m
5.023 12-20m m
5.03 Over burden drilling 115 to 133mm (in loose ground)
5.031 0-15m m
5.032 15-30m m
6.01 Supply, install and sheath MS 2'' dia, port distance 66cm m
grout sleeve pipes (tubes a
manschettes),
6.02 ditto in HDPE m
6.0 Supply, maintain, position Packers: all dia, all depths
and remove.
6.01 Single packer No
6.02 Double Packer No
6.03 MPSP-inlatable packer No
6.04 Circulation-packer incl return line No
6.05 Inlatable double-packer for tubes a manschettes No

Permeation grouting
7.0 Operating grout pump, incl Documentation of pressure,
weighing, batching, storing, rate and quantity as per EN
ducting of grout mixes of 12715 requirement
all kind, operating grout
pump under pressure,
electronic data aquisition
7.01 Hour operation of one only H
(irst) pump

173
Continued
Table 5.1 Proposed standard schedule of rates/ bill of quantities for grouting works (Continued)

174
Item

Ground improvement
Item No Group Main Sub Unit Quantity Price price
7.02 Hour operation of a second H
pump at same site location,
simultaneously with operating
the irst pump
7.03 Hour operation of a third H
pump at same site location,
simultaneously with operating
the irst and second pump
8.0 Material for grout mix
8.01 OPC (Blaine > 3.900 cm2/g) to
8.02 UFC (D80 < 12μm) to
8.03 Sodium-Bentonit to
8.04 Calcium-Bentonit to
8.05 Sodiumsilicat (liquid, 38ºBé) kg
8.06 PU (single shot) kg
8.07 PU (two-component mix) kg
8.09 Acrylat-(as per tenderers kg
proposal)
9.0 Borehole test Water pressure test in rock No
(Lugeon), permeability test in
loose ground (Lefranc), incl all
pumps, ducts, packers and
data recording/
documentation, all depth
Total, net Currency
 Permeation grouting 175

5.2 PREPARATORY WORKS AND DESIGN

It is important that the designer is aware of both the possibilities and the
limitations of grouting. The principles of luid mechanics might on irst
glance seem to govern grouting in the same way as they do the propaga-
tion of luids in other media like pipes and ducts. The lack of knowledge
of the intricate rheology of the grout and of the complex geometry of
low paths in the ground, however, poses serious problems to arriving at
mathematically ‘exact solutions’. Data on the rheology of the luid do lack
information on interstitial adhesion, surface tension relative to wetted
surface, and whether capillary forces would tend to support or prevent
penetration of the grout. It is particularly the small sections of intercon-
nected porosities for which it is dificult to develop a good low model.
This is because the very narrow parts of low channels will – particularly
when its diameter is getting close to the grain size of the suspension –
govern the penetration of grout more by actions of surface tension and
afinity of the luid relative to the wetted surface of the ground than by
iltration, viscosity and yield.
Cementitious grouts are the most common type of particulate grouts;
that is to say they contain particles (i.e., the grains of cement in a water
suspension). Cement as a material requires a water-cement ratio by weight
of about 0.38 to achieve complete hydration. However, in this form it would
be an extremely stiff paste, so for injection purposes additional water is be
added to the mix for the purpose of transporting the cement grains within
the issure (or pores of a sediment). The addition of water has the combined
effects of reducing the strength of the grout, increasing its shrinkage, and
increasing its setting time. The higher the water-cement ratio employed, the
weaker the grout, the greater the shrinkage, and the longer the setting time
that will result.
The question of setting time is important. Cements are manufactured
so that they have a setting time of about 4–5 hours. This period is stan-
dardised to provide a suitable period of workability for normal structural
applications. If we greatly dilute cements the setting time is delayed; 10–16
hours may result for water-cement ratios of 2:1 and 3:1, respectively. The
addition of clays or bentonite into the mix will further delay the setting
of the cement. Accelerator admixtures may be employed to reduce setting
times, but these work best on low water-cement ratio mixes and have the
disadvantage that they tend to increase the viscosity of the mix.
The penetrability of a cement-based grout into issures depends on two
main factors: the grain size of the cement used and the rheological (and
dispersive, particle-separating) properties of the suspension. As is well
known, the success of the grout is characterised by the size of the solid
particles of the grout in relation to those of the issures to be grouted.
However, to study the penetrability of a mix by merely studying the size
176 Ground improvement

of a single dry grain is misleading: single dry grains have a tendency

to grow in size during hydration and to agglomerate, thereby producing
‘locs’ larger than the single dry particle. Therefore, to improve the pen-
etrability of a particulate grout, it is necessary to both keep the grain size
low and reduce or prevent the tendency for single grains to locculate in
the mix.

5.2.1 Aspects of rheological laws for

particulate grout mixes
If we discuss penetrability we have to therefore discuss rheological prop-
erties of the mix. These are normally characterised by three parameters:
plastic viscosity, cohesion, and internal friction (surface tension which
strongly governs the penetration of ine issures, however, still remains
unresearched).
Figure 5.2 shows two laws of rheologic behaviour. Curve (1) is typical
of a purely viscous (Newtonian) luid. Water and many chemical grouts
such as silicates and acrylamides follow this law. Curve (2) represents the
behaviour of a so-called Bingham luid, which is characterised not only by
viscosity but also by cohesion. As discussed, cement grouts are not solu-
tions but particulate suspensions in water. If these suspensions are stable
(i.e., during grouting only a minor portion of water does become separated
from the cement) they do behave as a Bingham luid.
Assuming a stable, perfectly viscoplastic mix, Lombardi (1989) analysed
the low conditions of a mix through a smooth rock issure. He concluded

dv
1 τ = η dx
τ
dv
2 τ = C + ηB dx η′ dv
dx
2
η = Dynamic viscosity
ηB = Plastic viscosity
ηB 1
1 (dyn. visc. of plastic body)
η′B η′ = Apparent viscosity
C 1 η
1 dv C = Cohesion or yield value
dx
τ dv

dx

Figure 5.2 Rheogram of luids. (From Kasumeter, International Society for Rock
Mechanics, Widmann, R. (1996), Int. J. Rock Mech. Min. Sci. & Geomech.,
33(8):803–847.
 Permeation grouting 177

that the cohesion determines the maximum distance the grout can reach,
and the viscosity determines the low rate, and therefore the time necessary
to complete the injection at given pressures.
Simple theoretical considerations and elementary experimental evidence
show that, as soon as internal friction appears in a particulate mix, grouting
is no longer possible. If the cement grains are not transported by the luid
and come into contact, they will develop friction between the particles and
effectively stop grouting. This phenomenon is particularly important since
during the grouting process the initial water-cement ratio may decrease
due to loss of water under the applied pressure (i.e., pressure iltration) or
simply due to gravity (bleeding).
As we have seen from the previous discussion on grout theory, a low
viscosity grout mix is required to achieve suitable low rates in ine aper-
ture issures. The viscosity of the grout is normally measured by means
of the cone eflux test whereby the time required for a measured quantity
of grout to drain from a conical funnel of ixed dimensions is measured.
Various funnel geometries are used for different types of grout. For testing
common grouts used for injection purposes, the Marsh cone is normally
employed.
The ability of particulate grouts to penetrate into ine aperture issures is
controlled by the ability of the particles themselves to enter the issure and
by the degree of elastic widening of the issure that occurs during the grout-
ing process. Various authors have attempted to correlate these parameters.
The following comments indicate the range of results obtained.
The D95 particle size of a cement means that 95% of the particles within
the cement are smaller than this size. A Type I Ordinary Portland Cement
typically has a D95 particle size of about 60 microns, and a Type III rapid
hardening OPC (of the type commonly used for grouting) may typically
have a D95 particle size of about 40 microns. Microine cements are avail-
able with D95 particle sizes as small as 12 microns. The D95 particle size of
Bentonite clay is typically also about 60 microns.*
Following common iltration criteria, Karol (1990) like other authors
proposed for rock grouting that for cement-based grouts to penetrate a is-
sure within a rock mass, the aperture of the issure must be at least three
times the particle size of the cement grains. We consider that in practice
the joint width must be wider than this rule and that in addition to this
initial aperture, further elastic widening of the joint during injection is
necessary.
The ACEL research programme,† like other research (e.g., by Baban
1992) revealed (in carefully instrumented strain measurement tests using
microine cements and epoxy resin components) elastic opening and clos-

* A micron being a millionth part of a metre or one thousandth part of a millimetre.

† By Keil et al. of 1989 (quoted in Weaver and Bruce, 2007)
178 Ground improvement

ing of the fracture zone by as much as 100 microns during the grouting
operation. This dimension being some eight times the D95 particle size of
the microine cement decidedly proves the importance of the elastic defor-
mation of the rock mass for a successful treatment.
In attempted grouting of water bearing rock at a depth of 830 metres
using microine cement grouts, Naudts (1990) reported that it was not
possible to inject a stable microine cement grout into a strata observed
to have a permeability of over 100 Lugeons. This value would be con-
sistent with very heavily fractured rock (of >>10 fracs/meter and cor-
responding frac widths of 50 to 100 microns), which would normally
be expected to be readily treatable with such materials. In this case it
could be suspected that insuficient pressure might have contributed to
the phenomenon.This is because grouts, when entering a issure, have
to initially overcome resisting forces of friction, being frequently in the
range of speciied maximum grouting-pressure limits! (Feder, see FN,
ISRM 1996).
Suppliers’ catalogue information for microine cements indicates that for
the repair of concrete structures, microine cements with a D95 particle size
of 12 microns will penetrate cracks as ine as 0.25–0.3 mm. In this applica-
tion overpressurisation and elastic opening of the crack is not appropriate,
and this rule of thumb represents a ratio between crack aperture and par-
ticle size of some 20 times.
In order for cement grouts to be successfully injected it is necessary
for the cement particles to remain in suspension during injection. When
injected under pressure the mix may lose water into the issure. This loss
of water will cause a thickening of the mix and the generation of internal
friction, increased viscosity, and rigidity of the grout with (in the end) the
formation of a dense dry cake. These phenomena may eventually block any
further low of the grout into the issure.
If high grouting pressures are to be employed, the grout mix must have
a pressure iltration characteristic that is stable at the required pressure.
Pressure iltration testing requires either laboratory equipment like an API
ilter press or (for lesser requirements) may commonly be replaced by a
simple bleeding test, which measures the tendency for the cement particles
to settle out of suspension. In this test a fresh sample of grout is placed in a
glass measuring cylinder and covered to prevent evaporation. The amount
of free water left on top of the cylinder after two hours is generally termed
the bleeding characteristic of the grout. A grout is considered to be ‘stable’
if there is less than 5% of free water of the total volume of the grout sample
after two hours.
In order to produce stable grout (albeit with a relatively high viscosity)
without further additives or admixtures a water-cement ratio of about
0.6 to 0.8 is required for OPC cements and about 0.9 to 1 for microine
cements. The difference between the two would mostly arise from the
 Permeation grouting 179

iner particle size of the latter. However the rheological properties of these
grouts are unsuitable for normal injection applications, being too viscous
and cohesive.
Two courses of action are possible. It is possible to add additional water
to the grout mix to give suitably low values of viscosity and cohesion and
to stabilise the mix by the use of a colloidal additive. The most common
of these colloidal additives is bentonite clay, which is normally added to
the mix water in the ratio of 1%–6% by weight. When added to water
and hydrated, bentonite gives the mix water thixotropic gel properties, and
this acts to inhibit the settlement of the cement grains. However, the addi-
tion of bentonite strongly increases the cohesion, and to a lesser extent the
viscosity, of the mix. The bentonite particles are of a similar size to OPC
particles and do not affect the penetrability of these materials. However,
the bentonite particles are considerably larger than microine cement par-
ticles and stabilising microine cement grouts in this manner would limit
their performance. Other, more powerful, colloidal admixtures in the
form of long chain polymers such as methyl cellulose are available, but at
high concentrations these materials will signiicantly increase the cohesion
(reducing the reach of grouts) and viscosity of the mix. Catalogue informa-
tion indicates that a 2% solution of methyl cellulose in water at 20°C has
a viscosity of 4,000 centipoises. Note that the viscosity of water at 20°C is
approximately 1 centipoise.
The stability of grouts may also be improved to some degree by the
addition of very ine-grained pozzolanic materials such as silica fume.
This material is the by-product of the ferro-silica industry and generally
has to be imported in bulk from Norway. It therefore attracts substantial
transport costs for those outside Europe. The material is extremely ine,
having a grain size with a mean diameter of 0.1–0.15 microns, about 100
times iner than Ordinary Portland Cement. The material reacts with the
lime of the cement liberated during the hydration to form an amorphous
gel. Vipulanandan et al. (1992) reported tests on grout mixtures compris-
ing an Ordinary Portland Cement that had been stabilised by either the
addition of 5% bentonite, or 5% silica fume material. Both mixes had a
water/binder (cement + silica fume where used) of 1:1. He observed that the
measured bleeding of both mixes were similar, about 8% after two hours.
Penetrability, however (of the silica fume mix), would in comparison be
expected to be much better.
The alternative method is to use a stable grout mix with a lower water-
cement ratio and to achieve the required rheological properties by add-
ing of a super-plasticiser such as a Na-metacrylate. These admixtures are
surface active agents that negatively charge the cement grains and act to
reduce the grain agglomeration, apparent cohesion, and viscosity of the
grout. Grout mixes formulated with this type of admixture have the added
advantages that they are stronger, set faster, and suffer less shrinkage than
180 Ground improvement

conventional cement bentonite grouts. Set and workability are also easier to
control when using accelerator admixtures.
Stadler and Hornich (2008), for the purpose of design and prepara-
tory works, classify grouting into ive different categories. It should be
mentioned that all techniques associated with intended deformation while
grouting will not be considered in this section of the chapter. However, it
still is important to realise that every grouting application under pressure
is hydraulically introducing energy into the ground. Making hydraulic
forces act onto surfaces of grains in sediments, or onto the surface of is-
sures in rock, causes displacements even of a minor order be it intended
for deformation or not. So let us assume and accept that even permeation
grouting is a process where such (largely elastic, and only to a minor
extent plastic) deformations do occur, and in fact do support the pen-
etration of grout, the saturation of voids, and thus the success of the
treatment.
Design speciications for permeation grouting should carefully view this
aspect, particularly when limitations on grouting pressures are stipulated.
The difference between a useful and unavoidable (intrinsic) deformation by
the pressure in a propulsed grout low, and the avoidance of a frac pressure
which overpowers the structural resistance in the ground (causing undesir-
able deformation and heave) is not necessarily reconcilable with the weight
of the overburden! Today, this still remains the governing concept for the
stipulation of such a pressure limit. Viscous Bingham-luid grouts disperse
considerable energy in the irst decimetres after entering the ground, and
exhibit a strongly digressive pressure distribution in the porosities pene-
trated. Thus, the corresponding uplift forces remain limited. The ultimate
proof, however, still is a ield test to verify the genuine ground reaction
(instrumented with proper deformation gauges) under different pumping
velocities and grout types.
An overview on grouting techniques (principles and methods) is given
in Figure 5.3 and in EN 12715, Execution of Special Geotechnical Work,
Grouting (2001), under Pt 7.3.1.1. For the case of rocks, the Report on
Grouting (ISRM, 2000) is the literature of common reference. Ground
Improvement (Second Edition, 2005) contains a reference where a prac-
tical crossover between virgin permeabilities in soils and rock is com-
bined with types of grout mixes to choose, placement techniques, and
grouting protocols (for operative parameters like pressure, quantity,
energy, etc.).
Adequate knowledge of the relevant properties of the subsoil is of prime
importance for designing grouting works. It cannot be overemphasised
that with grouting being a predominantly hydraulic process, site investi-
gation does primarily require reconnaissance of hydraulic properties of
the ground. In particular, the stratiication of sediments and the type, fre-
quency, and orientation of discontinuities in rock are important features.
 Soil Rock
Type Fine Coarse Type Diffuse issurisation, Discrete joints (fracs)
Kakinites
Grout ways, Drillrods & Collapsble/ Stable rock Unstable
ports lances Unstable
Perforated pipes Ports Multiple packerSleevePipe
Sleeved manchette pipes single/double packer
Open-ended System Stage grouting Bottom Stage
pipes up
grouting
System Displacement (frac. compaction) Penetration (pore grouting, permeation) Frac(Fissure) Permeation grouting
grouting
Grout mixes Silicate/ Grout mixes Acrylate/
Acrylate Epoxy
Microine binder Microine binder
Bentonite/Cement Ordinary Portland Cement
Mortar Mortar

Virgin Kf 10.E-6 10.E-5 10.E-4 10.E-3 10.E-2 >>10.E-1 Virgin Lug. 1 5 10 25 50 >>100

Permeation grouting
Grouting Energy and Grout limited Limitation of Grouting Energy and saturation Pressurelimitation/Energy
parameters displacement quantities below quantity and parameters criteria criteria
criteria frac pressure until pressure split-spacing/from inside outwards//from outside inwards
resurgance or
interconnections
do occur

Figure 5.3 Overview of grouting techniques. (From Stadler, G. (2004). ‘Cement grouting,’ in Moseley, M.P. and Kirsch, K., Ground Improvement,
2nd ed Spon Press: Abington.)

181
182 Ground improvement

For grouting of soils, the following information is required:

r Stratiication, typical grain-size distributions, conductivity proile,

K h /Kv
r Porosity, saturation, speciic surface [m 2 /m3]
r Density of packing (CPT, SPT), grain shape, deformation modulus
r Mineralogical composition of the soil layers
r Groundwater table, gradient, GW-chemistry
r Position of wells, rivers, sewers, gullies, lines and ducts relative to
the intended grouting area; building foundations, basements, under-
ground structures and their respective conditions and properties adja-
cent to any intended treatment zone
r Soil pollution

The dimension of the measured unit of conductivity (permeability

coeficient) is m/s: This ‘velocity’, however, is related to the cross-section
of the ground as a whole. The ‘true velocity’ of a luid in the ground may
therefore be established only by relating the respective low rate (m3/sec)
to the available porosity.
In rocks, the following information is required:

r Lithological stratiication, stereo plot of discontinuities, transmissiv-

ity proile
r Frequency of discontinuities, modulus of deformation, porosity
r Anisotropy of transmissivity, RQD, mineralogical composition,
weathering
r Groundwater table, gradient, sources, barriers and wells, groundwa-
ter chemistry
r Position and conditions of any underground structures

Hydraulic testing in rock aims at quantifying the capacity of absorption

of water in litre per minute and per metre of borehole at 10 bar (excess)
pressure. The respective unit value is 1 Lugeon, which thus corresponds
to the volume-low of water at 1 l/min per linear meter of hole (irrespec-
tive of diameter, but 76 mm as a standard) at 10 bar pressure into rock.
The corresponding term is transmissivity (T) and the respective dimension
consequently is m 2 /s. Elastic or permanent deformations of rock, turbulent
low conditions, and so on may be identiied when interpreting test data
based on multiple pressure steps. Lugeon testing gives valuable information
on geotechnical and hydraulic conditions of the underground, but does not
necessarily relate to grout takes during later grout treatment because pen-
etration into issures and corresponding deformations do differ between
using water or grout.
 Permeation grouting 183

Since its early application at the Aswan project (built in the years 1960 to
1971, Figure 5.4), grouting of alluvial ground is accomplished by using the
sleeve pipe method (tube à manchette). Pipes inserted into boreholes usu-
ally are of 1½ to 2 inches in diameter (single-port 1/2-inch pipes are now
also used in uniform sands). The annular space between sleeve pipe and
borehole is sealed by a ‘plastic’ sheathing grout of cement—bentonites.
This sheath grout is intended to prevent grout escaping to the surface
instead of penetrating into the ground. However, this technique is not
suitable in rock.
Grouting through drill rods or driven pipes (lances) are techniques
for grouting applications of lesser requirement in coarse-grained ground
of high porosity and at low pressures. Grout mix in these cases will be
placed through perforated pipes which are driven or inserted in predrilled
boreholes; via borehole casings during withdrawal from the borehole; and
through the drill bit itself when drilling the grout hole.
The range of penetration and the degree of illing of voids using drill rods
or lances is limited. Fine-grained and cohesive soils are less apt to treatment
with particulate grouts or chemicals for reasons of iltration. Schulze (1993)
did some research on pore size distribution and penetrability of sediments
by relating grain size distribution of OPC and microine binders to the sieve
analysis of soil samples. Comparable efforts to deine the application of

Figure 5.4 Recent view of a cross-section of Aswan High Dam (on-site tourist poster).
184 Ground improvement

different grouts in soils lead to ‘groutability ratios’ commonly used in the

United States. There the D15 (diameter at which 15% of the soil sample is
passing) is related to a D85 (diameter at which 85% of the grouting material
is passing); according to this, groutability may be expected at ratios >24.
Also, success is not likely if grouting ine-grained sediments with a silt
content >5% when using particulate grout based on Portland Cement (PC)
which at 85% passing contains material of diameter >40 microns. For such
cases the use of microine binders is recommended instead. For application
where these microine suspensions cannot successfully penetrate, the only
remaining solution is to use chemical grouts.
Compaction grouting and frac grouting may be resorted to as a means to
consolidate or tighten the ground with soil-displacing methods of grouting.
Such systems make use of quite intensive pressures of 40 bar and higher.
Grouting in rock formations in the majority of cases aims at tightening
of issures against percolation of water. Groundwater that migrates and
lows under varying gradients in issures, joints, and tectonic discontinui-
ties under dams or in the form of seepage into deep tunnels will be reduced
or stopped by grout from such migration. Below 3 Lugeon, only microine
binders, acrylates or silicates (with organic hardeners) and sometimes res-
ins may be eficient.
To arrive at an assessment of likely average issure widths prevailing,
Cambefort (1964) explored relations between transmissivity (Lugeon
values), issure frequencies, and opening widths. In ISRM’s Report on
Grouting (2000) this approach is updated with more recent comparable
research and ield experience (see Figure 5.5).
Discontinuities in rock are dominantly two-dimensional in shape and,
moreover, are frequently intersected by other sets of issures and joints.
All of which neither appears plane and parallel nor is the opening width
constant. Consequently, the low regime in issures varies from ‘channel
low’ at low transmissivity (of <5 Lugeon), to concentrically ‘planar low’
starting off a singular intersection of a grout hole with a issure plane, and
inally to ‘spherical low,’ which activates a multitude of criss-cross issuri-
sations, suggesting quasi isotropic conditions. Flow equations are proposed
by Hässler and Gaisbauer (Widmann, 1993).
Grouting pressures generally drop exponentially with increasing distance
from the point of injection. Bingham luids tend to accentuate this pressure
drop compared to Newtonian luids, and because of this the problem of
issure widening or the danger of frac propagation is linked more to the use
of the latter.
Grouting open holes in stable rock is carried out in sections of 1.5 to
6.0m and from bottom up. Mixes for grouting in rock originally did make
use of unstable suspensions (separating >5% or even less free water in 2 h
under gravity) relying on the phenomena of pressure iltration for the suc-
cess of grouting. Nowadays, stable suspensions are preferred—and may
 Permeation grouting 185

Grouting in rock, assessment of fissure width

Chemicals Suspensions
ISRM (1996) 10–4 Acrylates - hand-gels – Micro – OPC

m
50/

10/
k[m/s]

Smooth joints
10–5

1/m
10–6

/m
b

5/m 10 2 50/m100
a c

/m 0/m

Rough joints
10–7 d
1/m

10–8

100 50 10 5 1 0.5 0.1 0.02 0.05 0.1 0.5 1

Transmissivity in [LU] Joint width [mm]
0.01

Number of fissures per metre (resp: efficiency)

(a) Isotropic rock
granular medium
(b) Orthotropic rock 0.1
(c) Smooth joints
n

(d) Rough joint

geo
Lu

1
on

ns
ns
01

uge

geo
geo
1L

Lu
Lu
100
10

10

100
Figure 5.5 Grouter’s diagram (From ISRM, modiied by Stadler, G., Howes and Chow
(1998). ‘100 years of engineering geology,’ symposium, TU-Wien).

also be easier produced than before. This is achieved by adding cement

additives or by using higher quality cement. However, a proper laboratory
testing procedure needs to be established and carried out.
Grouting to a predetermined refusal pressure has been abandoned as
well, in favour of a dual criterion in which grouted quantity and effective
186 Ground improvement

grouting pressure are multiplied (grouting intensity, which in essence cor-

responds to a limitation of grouting energy per m3 of treated ground) as
speciication for a general break-off criterion (Figure 5.6).
Sleeve pipes with inlatable jute bag packers (or multiple packer sleeve
pipes (MPSP), developed by Rodio around 1980) occasionally were suc-
cessfully used in collapsible rocks where the drilling to inal depth was
possible, either using casing or drilling muds.
When designing the borehole grid, making the choice of a grouting
method and specifying grouting parameters, it is essential to properly adapt
general ‘rules’ onto local geological geotechnical conditions, considering
the quantiiable aims of the treatment, topography and other limiting cir-
cumstances of the project.
In this context it must be noted that the reach of grout, and consequently
the borehole spacing, is inversely proportional to the yield value of the grout
mix. Reach of grout is increasing in proportion to the grouting pressures
applied. Viscosity of the grouting luid is responsible for frictional losses,
an aspect which must be considered when specifying maximum allowable
pressures or pressure losses, occurring when grout enters small issures
from a borehole.
Thus the maximum allowable pressure is deined ‘by itself’ and on the
basics of the hydraulic interaction of voids-geometry and rheology of
the grout mix rather than on the weight of the ground over the point of
injection.
It is evident that every grouting project needs expert preparatory action
and special attention to its design deinitions. Only measurable properties
should be addressed when specifying target values for grouting, and only
parameters which have a deined relation to grouted or ungrouted soils/
rocks should be selected. Visual inspection of grouted strata frequently fails
as a suitable criterion of acceptance. The designer obviously has to be an
expert. His realistic judgement of these questions will highly inluence the
outcome of the treatment.
Operative criteria were recently addressed by Semprich and Stadler
(2002) in a comprehensive way. The authors are of the opinion (in line
with requirements published in EN 12715) that the specifying of grouting
parameters has to already be dealt with in the design itself, even if—despite
of all improved theoretical background—this may still only be based on
experience and empirical data from past project realisations.
Whereas formerly the grout consumption alone (grouting rate and amount
at which a passé can be injected) determined the next steps (Weaver, 1993),
today it is a dual strategy which facilitates a more global assessment of the
proceedings. The maximum allowable grouting pressure frequently will be
ixed at around 80% of prevailing so-called frac pressure. This frac pres-
sure (at which the ground is separating and/or is losing its cohesive state)
may only be established by testing the ground at the individual project site
 Permeation grouting 187

using systematically stepped-up pumping rates. This methodical approach

makes the (guesswork) speciication of maximum pressures on the basis of
depth (in relation to the surcharge weight of the ground) obsolete. In allu-
vial soils, the respective allowable grouting pressures range between 5 and
35 bar with pumping rates varying between 5 and 15 l/min, respectively.
For issures in rock exhibiting <0.15mm width and using highly viscous
epoxies, these pressures might rise to as much as >120 bar without causing
any damage. The reason for this is that the pressure drop at the entry of the
issure is already consuming most of the destructive energy.
The speciication of a maximum quantity of grout to be injected per passé
or per unit volume of ground is based on the plausible estimate of accessible
porosity. Accordingly, for sediments these estimates vary between approxi-
mately 25% and 40%. For rock these quantitative limits are speciied—
frequently for economical than technical reasons—to prevent uncontrolled
loss of grout. Porosities in rock generally vary between 0.5% and 5%.
The grouting rate results from interactions between hydraulic frictions in
cross sections of porosities exposed to low depending on the rheology of
the luid (grout mix). Common applications of particulate suspensions are
operated at rates (as mentioned above) between 3 and 20 l/min. In karstic
rock this value might rise to even 100 l/min, or the limiting capacity of the
pump. Highly viscous epoxies, on the other hand, might have to be grouted
into issures of <0.15mm at rates of as low as <1 l/min.
In an effort to optimise extent and result of grouting works, a careful
monitoring of grouting data is recommended in EN 12715. There (among
others) the interpretation of Transient Pressure Data (TPA) and the limita-
tion of applied grouting energy (GIN, as the product of quantity of mix
grouted times grouting pressure, per linear metre of hole) provide new
diagnostic tools to the grouting process, which make it possible to quanti-
tatively discuss the applied grouting parameters against the original design.
Regarding adjustments to the grouting procedure and the recommended
steps that lead to the inal halt of the grouting operation, Weaver (1991)
formulated respective criteria that have been successfully applied in grout-
ing under dams and may be adapted to similar applications. His low charts
do supply the respective logic, indicating when to change rate, mix, or pres-
sures of grouting. Based on such or comparable considerations (TPA and
GIN), it becomes possible to formulate the design of modern grouting prac-
tice, particularly for grouting in rock.
An indication regarding relative costs for grout material may be drawn
from Table 5.2.
The deinition of operative parameters for penetration grouting of sedi-
ments is more dependent on the relation between geometry of pore sizes,
composition of particulate grouts, and rheology of the mix. Diagnostic
interpretation of the process during the grouting operation itself at pres-
ent remains limited. The interpretation of success or failure of grouting
188 Ground improvement

Table 5.2 Relative cost of grout material

Relative cost of diff. type
grout material, per kg
Types of grout material (provided but not injected)

Ordinary Portland Cement (OPC) 1

Binder 1–3
Microine Binder Blaine value 8,000 cm2/g 5
(MFC) Blaine value >12,000 cm2/g 10
Silicate gel (hardener: aluminate/acetate) 215
Resin products (e.g., polyurethane, specialised epoxies) >30–150

in distinctively orthotropic situations, as is the case with the stratigraphy

in most alluvial sediments, will therefore be even more dependent on the
relation between a K f horizontal and K f vertical than on an observation
of the development of the grouting pressures or rates, even an intricate
one. One of the methods practised is to either observe or interpret rates
at constant pressures, or pressures while keeping the rate of grouting at a
constant value.
The reality in tunnel grouting (as a modern application and revival
of grouting techniques), however, is that it is not possible to ‘design’ the
work with comparable precision in advance, which in many ways prohib-
its its comparability to this ‘design’ process. The design of tunnel grouting
operations is limited to the best estimates of the permeability and geom-
etry of issures in the rock through which the tunnel is to be driven, fre-
quently based on the average values only. Therefore, the basic design for
the grouting operation for tunnelling has to be reduced to an empirical,
observational basis.

5.3 EXECUTION OF WORKS

In general, grouting works should always be carried out by trained and

skilled personnel under competent and experienced supervision. Drilling
should make use of systems that least disturb the access for the grout into
subsoil porosities.
In spite of some drawbacks in terms of inluencing the size of pores and
issure intersections near the hole, roto-percussive systems are favoured—
mainly for economic reasons—and make use of external or down the hole
hammers, with or without casing, in rock as in alluvium. Rod size is nor-
mally 1¼ inches in diameter, and casings are up to 139 mm in diameter.
Direction and inclination of holes must follow the intentions of the
design. Two-percent deviation is normally an acceptable limit up to a depth
 Permeation grouting 189

of 20 m. However, it has to be kept in mind that horizontal holes and holes

drilled by percussion tend to delect more than others. Flushing of holes
with the aim to wash out ines or clayey materials from the ground has
limited effect and should in any case not be carried out at length.
The most effective way to fully provide grout into all underground voids
would be to address individually each and every issure, and each and every
individual stratum of sediment. Each of the porosities’ hydraulic proper-
ties could then be matched by the application of a rheological correspond-
ing mix, applied at optimum pressures, and supplied at optimum pumping
rates. However, this is neither technically feasible nor economically viable.
Therefore, an ‘averaging’ process is chosen as an economical compromise,
having the grouting ports installed at predetermined intervals (tube à man-
chette, TAMs)—irrespective of details in sedimentary stratiication—or (in
rocks) by separating individual borehole sections by packers at regular, uni-
form intervals, for example, at 1–6 m.
Different layouts and designs of grout-pipes and packers should therefore
be considered at the time when deciding the drilling method.

r Manchette pipes (TAM, tube à manchette); their undisputed advan-

tage is the reuse of the individual ports when grouting successive
phases using differing grouts.
r Single-port outlet mounted as a nonreturn valve at the bottom end of
a ½-inch pipe. This grout pipe may also be installed in bundles of sev-
eral individual supply lines, connecting to ports at different elevations
in the same hole, the advantage being that no manoeuvring of packers
is required when grouting at different depths.
r Multiple packer sleeve pipes (MPSP) do consist of a combination of
manchette pipes activated between jute bags inlated by cement grout.
Thus, even collapsible ground may be systematically treated in well-
deined sections.
r Open-ended or perforated lances driven into the ground by hammer
or hydraulics. These grouting devices provide access for grout in situ-
ations of lesser requirement, or lose ground exhibiting high conduc-
tivities. They are also limited in depth and installation accuracies.
r Single or double packers are used when grouting in rock or, the latter,
when grout is pumped into TAMs. Single packers set at the collar of
a hole in rock are frequently screw type expandable rubber packers
whereas, gas-inlatable single or double packers (between 0.3 and 1.5
m in length) may be lowered into holes as deep as 50 m. At greater
depth the risk increases of packers getting stuck and lost.
r Self-inlating rubber packers using the back-pressure of the grout
(being pumped through a nozzle in the packer or breaking through
a metal-membrane of deined bursting pressure) to inlate the sealing
element. This packer type may not be retrieved.
190 Ground improvement

Mixing of suspensions sounds like a trivial task; however, it is an art if

performed well to predetermined requirements. The requirements depend
on the task the grout has to fulil. For compaction grouting, the strength is
of minor interest, but the volume stability and expansion is important. For
grouting jobs in the tunnel environment, very often the strength criteria
combined with bleeding less than 1% and expansion in the range of 0.2%–
1%. Uniaxial compressive strength of up to 35 N/mm2 is nowadays stan-
dard. Depending on the water-cement ratios, bleeding and volume stability
become important issues, especially with ratios above 0.7.
Stationary plants using silos not only for cements but also for premixed
bentonites (for full hydration) and ine sands (in the case of using mortars
and pastes) do provide for suficient automatic functions to limit manpower
and increase capacity and accuracy. This helps to reduce mistakes and keep
to tolerances. In the last decade. major developments have been achieved
in terms of software developments and automation of supply plants. Fully
automated systems are not yet standard for all grouting jobs except large
operations, but this will probably change.
Stable mixes are now preferred and standard. This means that under
gravity no more than 5% free water should appear in a settlement test after
2 hours. However, for w/c-ratios below 1 using quality standard cement or
ready mixed binders bleeding in the range of 0.1% can easily be achieved.
Pressure iltration according to ASTM should not give more than 100 ml
of iltration water. Insuficient stability of the mix not only affects the
inal volume (lost by iltration), but also increases viscosity and yield, and
reduces setting time which reduces penetrability. To achieve a high quality
grout, conforming to design requirements, a proper mixing unit has to be
used. Fully automated mixers that do not only mix mechanically but also
circulate the grout are preferred. Modern mixer can rotate the grout up to
2,000 times per minute. In case a large volume of grout needs to be pre-
pared, agitators shall be used to keep the ready mixed grout in motion until
it is pumped to the injection point.
Grout pumps are mainly of the double-acting piston or reciprocative
plunger type. They are hydraulically driven and regulate any low rates
within the range of the capacity of the pump (usually 3 to 20 l/min). Pressures
may range up to 250 bar (for highly viscous epoxies); usually pumps should
be able to handle up to 100 bar at the corresponding minimum rate (i.e., 50
bar at 6 l/min, or 20 l/min at around 15 bar). The introduction of digital
hydraulic control management in the grouting pumps technique enables to
preset speciic values and injection parameters such as ‘pressure switch off
limits’, ‘delivery low rates’, ‘pressure prognostic curves’, and ‘GIN curves’.
Other pumps in use (for minor applications and standards) are sometimes
of the screw-feed type. Every pump is connected to a single grouting port
(packer position). Manifolds connecting more than one hole to a pump are
reducing the quality of the treatment.
 Permeation grouting 191

Table 5.3 Grouting strategies according to EN 12715 (CEN)

Rock
Stable Collapsible Soil
Lance,
Open borehole TAM Drillrod TAM casing
Single phase × × × ×
Multiple phase × ×
Bottom up × × × × × ×
Top down × × × ×

Pumps are connected to recording systems, which do help to follow pre-

determined quality assurance measures. In Table 5.3, some of the more
frequent strategies of grouting are presented, as published in European
Standards EN 12715.
In stable rock, it is common to drill the grout hole to the designed/
required depth and to start grouting in passes from bottom to top. Single
or double packer may be used. The use of single packer might result in reac-
tivating low of grout in the preceding pass. A new hole has to be drilled if
the same grouting area wants to be taken up a second time. Open boreholes
in collapsible rock are either treated top down (stage grouting, Table 5.3),
through TAMs or drillrods, or with multiple packer sleeve pipes. ‘Top
down’ stage grouting means that in a irst step the hole is drilled to a depth,
where the borehole walls still remain reasonably stable (but less than 6 m
to assure decent spread of grout penetration), a single packer is set at the
collar of the hole, and grout pumped into this irst section. TAMs will
only work if rocks permit suficient deformation for the sleeves to open;
therefore, MPS pipes are used, where the section between the jute packers
remains unsheathed by sealing grout, and is open for the cement grout to
spread and low into existing issures.
Another important consideration must concern not trapping water in the
pores or issures between already grouted areas, or to prevent grout from
escaping into areas where the treatment is not foreseen or to avoid grout
being lost outside the intended zone of treatment.
Prominent and typical examples for these kinds of problems is grouting
behind tunnel linings (Figure 5.7). Particularly if precompression of the
concrete lining is aimed at, it becomes of structural importance to avoid
anisotropic hydraulic loading.
A proper reporting system is recommended and, indeed, required to keep
track of operations and take adequate and timely decisions on the changes
to the procedures during the process. Electronic data acquisition is the stan-
dard today for the reporting of grouting parameters such as rate, quantity,
and pressures. Online transfer of these data may occasionally be arranged
192 Ground improvement

150

Resistance
to flow
(Shear strength
dynes/cm2)
Compressive strength (N/mm2)

50 100

40

30 28
Day 50
compressive 0.2
20 strength
Bleed
capacity
10 0.1

Bleed
0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Water/cement ratio (by weight)
1 kg.f = 981 × 103 dynes
Figure 5.6 Viscosity, shear strength, and bleeding of OPC suspensions. (From Littlejohn,
G. and Stadler, G. (1976). Joint lectures on anchoring and grouting at SAICE,
South Africa; and EUROCK 2004 & 53rd Geomechanics Colloquium,
Salzburg, Austria.)

even for remote control and interpretation. Storage and handover to the
engineer on discs for documentation purpose is standard. Interference in
the daily routines of a grouting operation from remote interpretation of
data is not recommended.
Quality assurance of the grout mix and its consistency are of prime
importance:

r Density is to be checked for correct content of solids (aerometer) or

on site using a scale
r Viscosity (at least Marsh low cone time in seconds, if not by a shea-
rometer). Be aware that low cone diameter are different in most of
the countries and not all low cones are suitable
 Permeation grouting 193

Figure 5.7 Drakensberg Pumped Storage Scheme, RSA, Tailrace Tunnel, Precompression
grouting, ESCOM, Rodio SA, 1970s.

r Yield (luid cohesion; Kasumeter, 2003; Heinz et al., 2003).

r Setting time (not at ambient but at ground temperature) of dehydrated
grout (remainder of an ASTM pressure ilter metre test), regular 250
ccm samples (including 28 day uniaxial strength), or of a ilm (adhe-
sive layer shed over a suitable base)
r Dispersion test (drop of mix squeezed between two 10 by 10 cm glass
plates and viewed under scaled magnifying-glass: 50×, or microscope,
against light).

5.4 MONITORING, CONTROLS, AND

ACCEPTANCE TESTS

The acquisition of grouting data makes an interpretation of the grouting

process possible, and conclusions may be drawn from these data as to the
success of the treatment.
194 Ground improvement

r Development of pressure and rate against time

r Hydraulic fracs or other respective movement of ground
r Interconnecting holes or so-called resurgences (spurt of grout to the
surface) and renards (French for larger ‘foxholes’ through which
grout may escape to surface) of grout.

Permeability tests before and after grouting may be helpful to assess the
degree of saturation achieved by grouting. The higher the virgin perme-
abilities of the ground, the greater the chance of a considerable improve-
ment. Wherever possible, upstream/downstream piezometers or seepage
rates should form the acceptance criteria for a successful impermeabilisa-
tion scheme and reduction of take of subsequent passes would indicate the
progress of illing.
Drilling energy measured when drilling (roller or ish-tail bit, borehole
supported by drilling mud) test holes before and after grouting indicate
the gain of strength. Core sampling for laboratory testing is more suit-
able in rocks than in soils. The system is inevitably failing in soils, even
using diamond core drills with uniaxial strengths of the cored material
dropping below 5 MPa, since obtaining intact samples, which would sat-
isfy laboratory requirements, is almost impossible. However, recovered
sample can still be used for visual inspection of the grouting success.
Open pit inspection sometimes is suficient to ascertain an improved
cohesion of grains or the visible presence of grout in the ground but
limited in depth.

5.5 RESOURCES AND EQUIPMENT

Drill rigs for the production of grout holes are of diesel/electrohydrau-

lic design with different mast conigurations and kinematics. The power
heads might provide (hollow stem if rods are passing through the power
head) spring-loaded or hydraulic jaws, or swivel-type connections (rods
only connected below a drill head) between driven rotating parts of the
head and the drill rods. The length of free travel on the mast is essen-
tial for productivity; every breaking and connecting manoeuvre of rod
couplings reduces production time for drilling. In Europe, percussion
and core drills by Atlas Copco, Casagrande, Huette, Wirth, and Klemm
produced respective machinery, which has satisfactorily performed over
decades.
Mixers and pumps are electrically driven and (as far as pumps are con-
cerned) have secondary hydraulic systems installed. Mixing may be cat-
egorised into mixing by agitation (paddle mixers, unstable mixes, batch
mixing), mixing by generating high shear forces (comparable to centrifugal
pumps), and mixing time, which should be limited to 30–120 s. Overlong
 Permeation grouting 195

mixing heats the grout and triggers the hardening process of hydration at
too early a stage. The mix should be kept (if at all) in agitating tanks where
low energy paddles keep the grout in motion and prevent particles from
sedimenting.
Many experts are of two minds about the requirements on the ‘evenness’
(continuity) of the grout low, though a slightly pulsating regime inds many
supporters. But what is even more essential is the possibility to regulate
low and pressure of the pump in a way that makes either a constant energy
concept possible or a constant rate or constant pressure scheme.
Pressure limiters with ON/OFF function are not suitable and hence not
recommended; pumps with an uncontrollable direct drive need a bypass
system which is prone to early wear; abrasive grouts should be handled
by low wear plunger pumps; ease of cleaning and maintenance is of great
importance (downtime!) especially when using chemicals.

5.6 GROUT MATERIAL

The correct choice of grouting material for the grouting works is of major
importance. In recent years, new chemical products have been introduced
on the market by several suppliers. Experience shows that the new materi-
als are often only tested under off-site ambient conditions and rarely under
conditions corresponding to the ones prevailing on site. Studying datasheets
may not be considered suficient. Depending on the grouting application,
laboratory tests need to be set up in accordance between designer and con-
tractor in an endeavour to appropriately test envisaged grouting materials.
The user needs to be aware that laboratory standards and experience may
be different in each country and even more different from continent to
continent.
Portland cement as a grouting material is well known and suitable for
most standard grouting applications. Very often in a two-stage grouting
process it can be found that in a primary grouting phase OPC is injected
with a Blaine value of around 3900 cm2/g or higher, and for the second
stage (to ill voids of a smaller cross-section) an ultra-ine cement (UPC)
may be applied. In Europe, the use of UPC is well established but on other
continents scarcely available. Also, a combined application of OPC fol-
lowed by a chemical grout such as a gel is common. Experience using foams
show that foams are not always suitable for permanent applications. Foams
should be used for tasks such as stoppage of water inlow into a tunnel
or into an excavation pit, and applying a second run using cement-based
grouts for ‘illing up’ or as a more rigid supplement.
In ‘modern grouting methods’ the requirement for controlled (low) vis-
cosity and yield (cohesion) may be achieved by adding of a ‘super-plasti-
ciser’ as an admixture, making a much lower water/cement ratio possible.
196 Ground improvement

The latter is recommended for making the grout more stable, less prone
to washout, producing less excess water under grouting gradients, and
avoiding pocketing of iltration water in the ground. The yield value (or
‘cohesion’ of the luid at zero low, measured by ball harp or Kasumeter)
should be carefully monitored to avoid producing ‘sticky’ grout, which
prevents free travel of the luid beyond a certain range.* Typically, when
using an OPC grout, the water/cement ratio would range from 0.8 to 1.5.
For microine cements (due to the greater ineness of these materials) the
water-cement ratio would be in the range of 1.1 to 2.0. A lower water-
cement ratio of any particulate grout makes it stable by its own constitu-
ents, and there is no requirement to add bentonites or other clay material.
The combination of low water/cement ratios and the absence of clay in the
mix has the advantage of

r The strength of the grout and its durability being high

r The shrinkage and permeability of the grout being reduced
r The normal setting times of the grout remaining maintained and, if
required, be controlled by the admixture of accelerators to the grout

Chemical grouts† typically are used as supplementary materials for spe-

cial situations and purposes, examples being where there are very strict
requirements in terms of permissible water or transgress, or where there is
need for the spontaneous local stoppage of percolating water low (Stephen
and Gert, 1999).
Where the indicated issure apertures (in rock) are assumed to be larger
than 0.2 to 0.3 mm the use of an OPC grout would still be appropriate.
In such cases the use of chemical grouts (such as silicate grouts) would not
only show poor economics but, may also be ineffective as a single-stage
grouting material. They could be ineffective because of shrinkage due to
syneresis (a chemical reaction depending on individual grout volumes), or
inability, due to low gel strength, to resist the higher water pressures at
depths. To be effective, chemical grouts should be used in a secondary or
tertiary grouting phase after the major issure structures have been illed by
stronger (and cheaper) cement or other particulate grouts.
A large number of chemical grouts are in fact available on the market
(Figure 5.8). Of the chemical grouts most commonly in use, silicate gels are
primarily mentioned (Hornich and Stadler, Grundbautaschenbuch, 2009)
of which the two main types are the so-called ‘hard gels’ and ‘soft gels’,
mainly differing in strength. With hard gels, strength of 1 to 10 N/mm2 can

* Remember that the formula (for the ‘range’ of grout travelling from a borehole) follows
principally: R=Pgrout*Awidth of void/τf yield of luid (grout) (where the dimensions would be for R
[m], P[bar], A [m], τ [bar]).
† Stephen and Gert (1999)
 Permeation grouting 197

Chemical reactive
components

Organic chemicals Inorganic reactives

Polymerisates Polycondensates Polyaddukte Inorganic gels

Unsaturated Urea-a. Phenol Epoxy resins Silica gels

polyester formaldehyde 1- and 2-component
acrylate resins resins polyurethanes
Tannin resins
Melamin resins

Organomineral resins
Silicate resins

Figure 5.8 Chemical Grouts, ISRM Commission on Rock Grouting, 1996.

be achieved. Soft gels are mainly used for sealing applications (for example,
in excavations) and do attain strength of 0.1 to 0.5 N/mm2. Gels consist
of approximately 50%–70% of water, 30%–45% of sodium silicate plus
hardener or locking agent. The major advantage of a hard gel compared
to cement mixes is the respective setting/hardening time. Hard gels may
develop their inal strength within hours. For a more detailed summary
on chemical grouts and special chemical grouts, the reader is referred to
Hornich and Stadler (2009) and subsequent paragraphs on silicate and
acrylate grouts.

5.7 CHARACTERISTICS AND APPLICABILITY

OF CHEMICAL GROUTS *

‘[Chemical grouts] is a generic term that can be applied to all forms of grout
that contain chemicals in solution either in water or with each other. The
family includes silicates, phenolic resins, lignosulphates, acrylamide, acry-
lates, acrylic, soluble lignates, sodium carbomethylcellulosis, amino resins,
polyurethane, polyester, epoxies, etc. Commonly these chemicals are dis-
solved in water to form aqueous solutions and rely on a chemical reaction
to cause a change of state from a luid to either a foam gel or a solid. In this
type of grout there are no suspended particles, hence by deinition chemical
grouts are stable and the application of a bleeding test to this class of grout

* Stephen and Gert (1999).

198 Ground improvement

is not required. Chemical grouts generally act as Newtonian luids; display-

ing viscosity but not (or only very low) values of cohesion. Consequently
there are no absolute limits on penetration and this factor is limited only by
practical or economical considerations of acceptable grouting/gel times and
(to a marginal extent) on pressure limitations.
The principal application of chemical grouts is in the grouting of soils
rather than rock. The majority of products commercially available and eco-
nomically acceptable are formulated for this market and these materials,
with some exceptions, will generally lack suficient strength and durability
for applications involving stemming the low through open issures in rock
and when subjected to high hydrostatic pressures.
When reviewing the literature, one may ind that silicate grouts are
used in conjunction with a pre-injection of cements stabilised by ben-
tonites. Misleadingly, the term ‘chemical grout’ has become to a certain
extent (and in certain areas of the world) synonymous with the use of sili-
cate (and acrylamide) grouts only. These grouts therefore are dealt with
in some detail.

5.7.1 Silicate grouts

Sodium silicate grouts have been extensively used in urban areas for more
than 15 metro schemes, including those in London, Paris, Vienna, Cairo,
Caracas, and Hong Kong. They consist of liquid silicate plus water and
hardener, are generally considered to be nontoxic, and there have been
no incidents of signiicant pollution or environmental damage recorded,
although unpleasant smells have been reported associated with the use
of particular hardeners during the construction of Auber station on the
French RER project. In that very case, the reagent used was ethyl acetate
(which is now out of use).
Silicates used for grouting are usually manufactured by fusing a mixture
of silica sand and sodium carbonate at 1400°C in a furnace. The vitre-
ous silicate obtained is subsequently dissolved in water under pressure at
a temperature of 150°C to give a syrupy liquid: Liquid silicate or ‘water
glass’. The liquid silicate in its concentrated form (38° Beaumé) is too vis-
cous for direct injection and may have a viscosity in the range of 40 to 180
centipoises at 20°C depending upon balance between the silica and sodium
molecules in the formulation. The viscosity of water at similar temperature
is approximately 1 centipoise.
For grouting purposes, it is necessary to dilute the liquid silicate with
water. The greater the dilution, the lower the viscosity of the grout and
the greater the low rate that may be achieved. However, the greater the
dilution, the weaker the gel strength of the set grout and the greater the
tendency to syneresis problems. To achieve a viscosity of ive centipoises
(as speciied) implies a high dilution, low strength gel of the type typically
 Permeation grouting 199

employed for injection into completely weathered rocks and soils after hav-
ing been pretreated with particular suspensions.
Liquid sodium silicates are highly alkaline, with a pH in the order of 10.5
to 11.5. They will react with an acid or acid salt to form a gel. Amongst
the gelling agents commonly used were sodium bicarbonate, sodium alu-
minate, and various other inorganic or organic acids. In the late 1950s a
new generation of gelling agents was developed consisting of methyl and/
or ethyl diesters formed from the action of aliphatic diacide mixtures on
methanol and/or ethanol. These gelling agents when dispersed in a sodium
silicate solution in the correct proportions go through a slow saponiica-
tion, which after a predetermined time provokes the liquid to gel in the
form of a white mass: SILICA GEL. These reagents are proprietary chemi-
cal systems, such as Hardener 600, and are marketed by specialist com-
panies such as Rhone Poulenc of France. The silicate grouts in common
usage today are a combination of a sodium silicate resin with a proprietary
chemical agent (hardener).
In application for grouting, the gelling agent or hardener is mixed with
the diluted sodium silicate shortly before injection into the soil. This grout
penetrates into the interstitial voids between the soil particles, conferring
on the formation the required cohesion and impermeability when solidify-
ing during setting.
Silica gels suffer from the phenomena of syneresis, which refers to the
progressive extrusion by the gel of a signiicant quantity of water. The
phenomenon is particularly problematic in dilute low viscosity gels, which
have been formulated for a long setting time. When injected into ine sands
syneresis is not normally a problem when the interstitial pore dimensions
within the soil are small. However, the phenomenon may become problem-
atic if injection takes place into coarse sands or open issures in rock where
the pore aperture dimensions are larger. In these materials the strength
of the gel itself becomes of importance and syneresis can lead to the inal
failure of the gel.
Silica gel can also be subject to washing out. When grouted sand samples
are immersed in lowing water it is observed that the gel will progressively
break down, leading eventually to a complete disintegration of the sample
for certain gel types. The water solubility of the silica gel is due to the
presence of non-neutralised soda, which attacks the silica. Laboratory tests
have shown that the higher the non-neutralised soda concentration, the
greater this solubility becomes. Shirlaw (1987) described the piping failure
in coarse beach and alluvial sands treated with silicate grout and attributed
the failure to these mechanisms. The same formulations had previously
been successfully when injected into ine-grained sands and weathered
rocks that had a smaller pore structure.
It should be noted that the suppliers of gelling agents for sodium silicate
grouts do not recommend these materials for grouting of issures in rock,
200 Ground improvement

preferring to recommend acrylamide-based materials which do produce a

stronger and more stable gel for these applications.

5.7.2 Acrylamide grouts

This type of grout has been used successfully in various rock grouting
projects where very low permeability results were required to be achieved.
Acrylamide-based grouts consist of a mixture of two organic acrylamide
monomers, which forms between 90%–97% of the mixture, and a cross-
linking agent such as methylene-bis-acrylamide that forms the balance. The
higher the percentage of the cross-linking agent, the stronger the resultant
gel. Grout solutions up to 20% solid have viscosities less than 2 centipoise
and are readily injectable into very ine issure structures—but only at fairly
low pressures! (Remember: ‘injectability’ of chemicals depends more on
structural properties of the luid (chemical-molecular chain-length, size
and complexity, surface tension and cohesion), whereas viscosity mainly
governs friction of low and thus, grouting pressures applied—but not so
much penetrability itself.) Such solutions when properly catalysed will,
after a length of time dependent upon the catalyst concentration, change
almost instantly into a solid, irreversible gel.
The principal dificulties with these materials are their very high cost and
the potential toxicity of the components. Acrylamide grouts are (depending
on make and speciic concentration) a potentially neurotoxic poison and as
such may represent a considerable hazard to the operatives employed in their
use. In practical applications it is not always possible to ensure complete
neutralisation of the grout and it may not always be possible to prevent grout
from leaching into local watercourses, leading to acrylamide poisoning.
Acrylamide grouts were irst introduced in the United States in 1953 but due
to the toxicity problems were withdrawn in 1978. A similar product, Nitto
SS, was withdrawn in Japan following a careless application near a well,
which led to several cases of acrylamide poisoning. A similar French prod-
uct, Rocagil BT, was used with success on the Hallendsas project in Sweden
but had to be withdrawn after the material leached into a local stream and
poisoned livestock. Neurological problems associated with loss of sensation
and motor control in limbs were also reported from this project.’

5.7.3 Grouts having to fulfil

environmental standards *
Chemical grouts have to be formulated and applied in a way so as not to create
undue hazards or transgression of environmental standards. In this respect
two types of product descriptions of grouting materials may be relevant.

* Stadler (2001).
 Permeation grouting 201

First, the new Eurocode on grouting EN 12715 of 2000, which states:

6.2.5 Chemical products and additives:

6.2.5.1 Chemical products such as silicates and their reagents, lig-
nin based materials, acrylic or epoxy resins, polyurethanes or others
can be used in grouting work subject to compliance with environmen-
tal legislation.
6.2.5.2 The effects of all products and by-products resulting from
reaction of the chemical products with other components of the grout
or with the surrounding ground shall be considered.
6.2.5.3 Admixtures are organic or inorganic products added in
small quantities during the mixing process in order to modify the
properties of the grout and to control the grout parameters such as
viscosity, setting time, stability, and strength, resistance, cohesion and
permeability after placement.
6.2.5.4 Admixtures to grout such as super plasticisers, water retain-
ing agents, air entrainers and others are subject of parts 1, 3, 4 and 6
of prEN 934 and prEN 480-1 to 480-12.

Second, refer to legislation in relation to the term ‘toxicity’. In fact, many

international standards do differ considerably on this subject.
Relevant standards do require that any

…environmental impact, particularly the toxicity of the grout and the

grout components and their effect on the ground and drinking water
should be considered before grouting. When testing the grouting mate-
rial for environmental impact, the following aspects should be consid-
ered: (a) whether during processing, transport or grouting, substances
can be generated or released which could be hazardous to the environ-
ment or the grouting crew; (b) whether noxious substances can spread
upon mixing with groundwater; (c) whether reaction products can be
produced or released which inluence the water quality; (d) the type
of particles eroded from the hardened grout; (e) chemical reactions
between hardened grout and groundwater.

It seems, however, that no oficial/reliable deinition of the term ‘toxic’

does exist, and frequently codes like the Canadian Environmental Protection
Compendium may have to be consulted, which states, ‘The term “toxic”
refers to the ability of a physical, biological or chemical agent to provoke
an adverse effect or deleterious response in an organism.’ The compendium
further notes that: ‘… some jurisdiction extends the term organism to the
environment as a whole.’
Unfortunately it is these words: ‘…harmful, adverse, deleterious, nox-
ious, hazardous, dangerous, negative, lethal…’ which may leave engineers
202 Ground improvement

with a semantic confusion when e.g. preparing the wording of a construc-

tion speciication, be it for temporary or permanent grouting measures.
The toxicity of a substance is commonly measured in terms of the
effects caused by oral ingestion. However, the wider deinition of the
term also includes adverse effects such as dermal irritation through skin
contact or damage to eyes and the respiratory system. Such reactions are
of no less concern, although toxicity levels for such reactions are poorly
deined. Where such reactions are of concern, standards such as the EEC
directives concerning labelling conservatively enforce safety procedures
based simply on the classiication of a substance but irrespective of its
concentration.
The oral toxicity of substances is tested on laboratory animals. The lethal
single oral dose of the material that will kill 50% of the sample population
is termed the LD50 and is quoted in terms of milligrams of the substance per
kilogram of body weight of the animal.
Various classiication systems existing for toxicity and the principal sys-
tems are shown here:.

UK—EEC
Very toxic LD50 = 0–25 mg/kg
Toxic LD50 = 25–200 mg/kg
Harmful LD50 = 200–2000mg/kg
USA
Very toxic LD50 = 5–50 mg/kg
Moderately toxic LD50 = 50–500 mg/kg
Very slightly toxic LD50 = 500–5000mg/kg
Canadian Environmental Protection Compendium, in an inverse order
to the above:
Practically non toxic LD50 = > –15,000 mg/kg
Slightly toxic LD50 = 5000–15,000 mg/kg
Moderately toxic LD50 = 500–5000 mg/kg
Very toxic LD50 = 50–500 mg/kg
Extremely toxic LD50 = 5–50 mg/kg
Super toxic LD50 = <5mg/kg

Karol describes the LD50 for acrylamides (used in acrylamide grout prod-
ucts such as AM9, Nitto SS, AV100, etc.) to be 200 mg/kg and the LD50 for
the methylene-bis-acrylamide commonly used as the cross-linking agent as
390 mg/kg. Karol quotes the methanol acrylamide used in ROCAGIL BT
as having a toxicity of 50% of Nitto SS.
Obviously and in consequence, when working with chemicals which to
any principal extent might be toxic, it is necessary to establish safe expo-
sure levels signiicantly below the fatal dose and to ascertain any degree of
cumulative toxicity.
 Permeation grouting 203

Many and most of the ‘adverse effects’ (from skin irritation to neurotox-
icity) are not fatal, however, but obviously do already occur at consider-
ably lower levels of exposure than the LD50 dosage. In summary, the safe
application of chemical grouts has to be meticulously established by proper
planning before any use on site. Grouts which require extensive quality
assurance programmes (as a consequence of differentiated dosing and mix-
ing programmes on site) should be used last. Grouts which are particularly
reaction-sensitive to even marginal dosing errors and temperatures should
be avoided. The fact that ever-demanding speciications are aiming at per-
meability coeficients below 10 -7 m/s in sandy gravels or below 0.1 Lugeon
in rock will increasingly make the use of chemicals beyond the range of
application of microine binders inevitable in future.
The more questions of material interrelate with aspects of engineering
application, the more it becomes the duty of engineers to perceive both
disciplines when realising demanding grouting works. They also have to
be prepared and educated enough to accept responsibility not only for the
engineering aspects, but for questions of safety and the proper environmen-
tally responsible use of such materials.’
For special grouting applications where high strength is a criterion,
cement-based grouts need to be modiied with additives to fulil design cri-
teria such as volume stability, degree of expansion, and degree of sedimen-
tation. Quite a number of additives are available on the market. The user
needs to be aware that the majority of the additives have been developed
for the concrete industry and just been modiied to be used in grouts. It is
important to undertake proper laboratory testing to verify the suitability
and successful functioning of the additive. The average volume of additives
used in cement-based grouts is in the range of 1%–2%. The laboratory tests
shall also consider the way grout will be mixed on site, considering the rota-
tion per minute, the time of mixing, and the order of adding the materials
into the mixer. The latter is particularly important when mixing chemical
grouts. Chemical grouts such as resin or gels do often consist of two to four
components. For small grouting jobs this is suitable; however, for larger jobs
with large quantities the appropriate equipment for preparing a grout out of
four components may often not be available. As mentioned before, cement is
the most common material for producing grout, especially Portland cement.
Blended hydraulic cements, blast furnace cements, and other special cements
such as microine cement are less used for grouting. When designing a grout-
ing application, it is important to be aware that cement standards and dei-
nitions differ between countries. For example, in Europe cement is deined
by its strength followed by its mineralogical composition and original prime
material. In North America, cement is basically deined by its application,
originating materials, and its characteristics of strength development. In
recent years, cement manufacturers have specialised and developed ready-
mixes of cement-based material (so-called binders) for standard grouting
204 Ground improvement

applications and also for illing the annular space between the manchette
tubes and the ground (sheath grout). One of the popular standard additives
to grouting mixes is clay containing high proportions of montmorillonite.
In the construction industry these clays are better known as bentonites.
Bentonite is usually added to reduce the sedimentation of aggregates (pro-
viding for ‘stability’ of the mix), and does change the low characteristic
(rheology) and viscosity of the mixed grout. This can be achieved by adding
just 1 to 2% of bentonite (by weight of cement).

5.8 TECHNICAL SUMMARY

Literature over the last few decades has seen quite a number of reports
on successful grouting applications, with detailed reports about their
design, execution, and performance, including many hydro and irrigation
dams (ICOLD). Several useful conclusions may be drawn from these expe-
riences, but it is not always wise to compare project situations without
detailed knowledge on ground conditions and drilling-grouting technolo-
gies applied, or targets set and (measurably) achieved. In order for cement
grouts to be successfully injected, it is necessary for the cement particles
to remain in suspension during injection. Equally important are effective
grouting pressures, suficiently high to overcome substantial pressure losses
when entering ine voids and to enlarge issures elastically in order to facili-
tate the entry of the grout particles.
Simple theoretical considerations and elementary experimental evidence
show that, as soon as internal friction in a particulate mix occurs, grout-
ing is no longer possible. The penetrability of a cement-based grout into
issures depends on two main factors: the grain size of the cement used and
the rheological properties of the suspension.
However, merely studying the size of a single dry grain is misleading: sin-
gle dry grains have a tendency to grow in size during hydration and agglom-
erate, producing ‘locs’ larger than the single dry particle. To improve the
penetrability of a particulate grout (suspensions are most popular because
of being cheap), it is necessary to both keep the grain size low, stay within
or delay the start of hydration, and reduce/prevent the tendency for single
grains to locculate in the mix.
The question of setting time (irrespective of the type of grout) is impor-
tant for the management of the grouting process against time, and the
choice of a correct treatment system altogether. Cements are manufactured
so that they have a setting time for industrial applications of about 4–5 h. If we
greatly dilute cements the setting time is irst delayed (10–16 h may result
for water-cement ratios of 2:1 and 3:1, respectively), and then accelerated
again during iltration. The addition of clays, bentonites, or accelerator
admixtures reduces setting times (simultaneously increasing the viscosity
 Permeation grouting 205

of the mix). It is quite clear that the rheological behaviour of the suspension
follows delicate relationships which have to be monitored and engineered
on a continuous basis.
In conclusion, the essential ingredients for a successful grouting proj-
ect are

r To go about any grouting project as open (educated) and engineering-

minded as possible
r To perform under continuous questioning/reafirming of the geotech-
nical model of the ground (in partnership with the designer - prefer-
ably an experienced Engineer or an Engineering geologist)
r Under permanent perception of the phenomena observed and inter-
pretations derived from these (possibly cross-checking these with an
experienced grouting foreman)

REFERENCES

Akjinrogunde, A. (1999). ‘Propagation of cement grout in rock discontinuities under

injection conditions,’ Institute for Geotechnics, University of Stuttgart, Vol. 46.
Baban, O.R. (1992). ‘Crack-injection in to massive concrete with synthetic resin,’
PhD thesis, Institute of Materials Sciences, Faculty of Civil Engineering, Graz
Technical University.
Baker, W. Hayward (ed.) (1982). ‘Grouting in Geotechnical Engineering,’ Conference
Proceedings, New Orleans.
BS EN 12715. (2000). ‘Execution of special geotechnical work. Grouting.’ July.
Cambefort, I. (1964). Injections des sols, Eyrolles, Paris.
Doran S. and Stadler G., unpublished, generalised expertise on chemical grouting,
Hong Kong, Nov. 1999.
Ewert, F.-K. (1985). Rock Grouting with Emphasis on Dam Sites, Springer Verlag:
Berlin–Heidelberg.
Heinz, A., Hermanns Stengele, R., and Plötze, M. (2003). ‘How to measure rheo-
logical properties of bentonite suspensions on construction sites,’ Annual
Transactions of the Nordic Rheology Society, Vol. 11.
Houlsby, D. (1989). ‘Cement Grouting,’ Dep. of Water Affairs, Sydney, Australia.
Hornich, W. and Stadler, G. (2009). Grundbau-Taschenbuch, 7. Aulage, Teil 2,
Kapitel 2.3 ‘Injektionen,’ Ernst & Sohn: Berlin.
Karol, Reuben, H. (1990). Chemical Grouting, 2nd ed., Marcel Dekker: New York.
Kasumeter, International Society for Rock Mechanics, Widmann, R. (1996),
Commission on Rock Grouting, Int. J. Rock Mech. Min. Sci & Geomech.,
33(8):803–847.
Kasumeter, (2003), ‘How to measure rheological properties of Bentonite suspensions
on construction sites’, Heinz, Hermanns Stengele, Annual Transactions of the
Nordic Rheology Society, 11.
Keil et al. (1989). Quoted in Weaver & Bruce, ‘Dam foundation grouting,’ ASCE
Publications, 2007.
206 Ground improvement

Littlejohn, G. and Stadler, G. Joint lectures on anchoring and grouting at SAICE,

1976, South Africa; and EUROCK 2004 & 53rd Geomechanics Colloquium,
Salzburg, Austria.
Lombardi, G. (1989). ‘The role of cohesion in grouting,’ ICOLD Proceedings,
Lausanne.
Naudts, A. (1990). ‘Revolutionary changes in the grouting industry resulting from
polyurethane injection technology,’ Second Canadian International Grouting
Conference, Toronto.
Nonveiller, E. (1989). Grouting, Theory und Practice, Elsevier: Amsterdam.
Schulze, B. (1993). Neuere Untersuchungen über die Injizierbarkeit von
Feinstbindemittel-Suspensionen. Berichte der Int. Konf. betr. Injektionen in
Fels- und Beton, A.A. Balkema: Rotterdam, 107–116.
Semprich, S. and Stadler, G. (2002). ‘Grouting.’ In: U. Smoltczyk (ed.), Geotechnical
Engineering, Geotechnical Handbook, Ernst & Sohn: Berlin.
Shroff, A.V. and Shah, D.L. (1993). Grouting Technology in Tunnelling and Dam
Construction, A.A. Balkema: Rotterdam.
Stadler, G. (1992). Transient Pressure Analysis of RODUR Epoxy Grouting at
Koelnbrein Dam, Austria, Diss. Thesis, MUL, Leoben.
Stadler, G. (2001). Seminar on waterprooing of tunnels, CUC-Seminars, F. Amberg,
Sargans, Switzerland.
Stadler, G. (2001). Permeation grouting, ASCE Seminar Publications: New York.
Stadler, G. (2002). ‘Was hat die internationale Normung der Injektionstechnik
gebracht,’ Injektionen in Boden und Fels, Ch. Veder Colloquium, Graz
University of Technology, pp. 1–20, April.
Stadler, G. et al. (1989). ‘Pressure sensitive grouting,’ Technical University, Vienna,
Stadler, G., Howes and Chow (1998). ‘100 years of engineering geology,’ sympo-
sium, TU-Wien.
Stadler, G. (2004). ‘Cement grouting,’ in Moseley, M.P. and Kirsch, K., Ground
Improvement, 2nd ed. Spon Press: Abington
Stephen, D. and Stadler, G., unpublished, generalized expertise on Chemical Grouting,
Hong Kong, Nov. 1999.
Verfel, J. (1989). Rock Grouting and Diaphragm Wall Construction, Elsevier:
Amsterdam.
Vipulanandan, C. and Shenoy, S. (1992). ‘Properties of cement grouts and grouted
sands with additives,’ Proceedings, ASCE Specialty Conference on Grouting,
Soil Improvement and Geosynthetics, pp. 500–511.
Weaver, K. (1991, 2007). Dam Foundation Grouting. American Society of Civil
Engineers: New York.
Weaver, K. (1993). ‘Selecting of grout mixes – some examples from US practice,’
Proceedings of Grouting in Rock and Concrete, A.A. Balkema: Rotterdam,
pp. 211–218.
Widmann, R. (ed.) (1993). Grouting in Rock and Concrete, A.A. Balkema:
Rotterdam.
Chapter 6

Jet grouting
George Burke and Hiroshi Yoshida

CONTENTS

6.1 Introduction ................................................................................. 208

6.2 History ......................................................................................... 209
6.3 Theory of jet grouting................................................................... 213
6.3.1 Effect of dynamic pressure................................................. 213
6.3.2 Effect of low rate .............................................................. 216
6.3.3 Effect of compressed air..................................................... 217
6.3.3.1 Effect of compressed air shrouding....................... 218
6.3.3.2 The velocity and volume of compressed air .......... 218
6.3.4 Effect of the soil................................................................. 219
6.3.5 Other effects ...................................................................... 220
6.3.6 Practical considerations ..................................................... 222
6.3.6.1 Design parameters for jet grout material .............. 223
6.3.6.2 Drilling tolerances ................................................ 226
6.3.6.3 Control of jet grout returns .................................. 227
6.3.6.4 Sequence of construction ...................................... 228
6.3.6.5 Quality control and validation ............................. 228
6.4 Application of jet grouting............................................................ 229
6.4.1 Groundwater control ......................................................... 230
6.4.2 Underpinning..................................................................... 232
6.4.3 Tunnels and shafts ............................................................. 233
6.4.3.1 Bottom slab .......................................................... 233
6.4.3.2 Subsurface props .................................................. 234
6.4.3.3 Roof barriers ........................................................ 235
6.4.4 Environmental applications ............................................... 237
6.4.4.1 Encapsulation ....................................................... 237
6.4.4.2 Containment barriers ........................................... 238
6.4.4.3 Active barriers ...................................................... 239
6.4.5 Waterfront structures......................................................... 241
6.5 Case histories................................................................................ 242

207
208 Ground improvement

6.5.1 Seismic remediation: Wickiup Dam, La Pine, Oregon,

United States ...................................................................... 242
6.5.2 Compression ring access shaft: Grand & Bates Sewer
Relief, St. Louis, Missouri, United States ........................... 244
6.5.3 Groundwater control: Waste-Water treatment Plant,
Providence, Rhode Island, United States............................ 248
6.5.4 Underpinning: Vassar College, Poughkeepsie, New
York, United States ............................................................ 250
6.5.5 Environmental: Philadelphia Airport, Runway 8-26,
Pennsylvania, United States ............................................... 251
6.5.6 Settlement control: Japan................................................... 252
6.5.7 Cofferdam sealing: New Orleans, Louisiana, United States....252
6.5.8 Waterfront structure: Battery Park City Authority, New
York City, United States ..................................................... 253
References ............................................................................................. 256

6.1 INTRODUCTION

Of all forms of ground improvement systems, jet grouting must be regarded

as one of the most versatile. With this technique it is possible to strengthen
in-situ soils, cut off groundwater, and provide structural rigidity with a sin-
gle application. In particular, jet grouting can create the highest-strength
treated ground (soilcrete) of all the ground improvement systems. It can
also be regarded as one of the most technically demanding of ground
improvement systems requiring both technical excellence in design and
construction because failure of either component will result in failure of
the product.
Figure 6.1 shows the principal method of application whereby either
high-pressure water or grout is used to physically disrupt the ground, in
the process modifying it and thereby improving it. In normal operation the
drill string is advanced to the required depth and then high-pressure water
or grout is introduced while withdrawing the rods.
As discussed in Section 6.2, jet grouting has a long history of develop-
ment from its initial use to current practice. In the ield of jet grouting, the
most notable advancements have been in Japan where the technique has
been reined to its present-day capability by careful attention to detail in all
aspects of the system. Through the years, careful research and execution
has resulted in increasing column diameter and range of applicable soils.
This development is also set out in Section 6.2.
After reading this chapter, it is hoped that the practicing engineer will
understand how jet grouting came into existence, the technical complexity
and design requirements needed for a successful application, and the range
of applications for which jet grouting can be used.
 Jet grouting 209

Figure 6.1 Jet grout column construction.

6.2 HISTORY

The scouring power of water has probably been employed as a soil excava-
tion method since early times, especially in the mining industry, where use
of it is documented in the Middle Ages.
The earliest patent regarding jet grouting was applied for in England
in the 1950s; however, the real practical development of jet grouting took
place for the irst time in Japan. This technology was initially aimed at
improving the effectiveness of water tightness in chemical grouting by erod-
ing the untreated or partially treated soil, which was then ejected to the
surface for disposal, being replaced with cement-based slurry for impervi-
ousness. Subsequently, jet grouting was irst applied to create thin cut-off
walls, as shown in Figure 6.2.
For preventing water ingress, a derivative of panel jet grouting was
evolved which sealed the gap between declutched sheet piles, for example.

Figure 6.2 Exposed jet grout panels.

210 Ground improvement

Figure 6.3 Jet grout sealing between piles.

This derivative allowed the formation of part columns (shown in Figure

6.3) by causing a twin-angled jetting motion or a windscreen wiper motion
of the monitor during lifting.
In the early 1970s, rotating jet grouting emerged in Japan because of the
fact that panel jet grouting could hardly create satisfactory products due to
varying thickness and somewhat fragile strength.
In the mid-1970s, jet grouting was exported to Europe and since then has
become popular worldwide. According to required geometry, three main
variants of jet grouting have emerged in the same period, of which concep-
tual schematics are illustrated in Figure 6.4. One of the variants is called
the single system (S), which is the simplest form of jet grouting, ejecting a
luid grout to erode and mix with the soil.
In certain soil types spoil can be more viscous and, without the aid of an
airlift, cannot easily travel up to the surface and heave may consequently
occur. In cases where heave would cause serious damage (e.g., when under-
pinning a building), 100% relief through the annulus of the borehole has
to be guaranteed. When drilling signiicantly below the groundwater level,
eroding effectiveness can be considerably reduced on account of the absence
of the shrouded air, which increases cutting energy.
The double system (D) adds compressed air, which surrounds (shrouds)
the grout jet to enhance the erosive effect, especially below the water table.

Air
Grout Grout Air
Water Air
Air Grout
Air
Air
Grout

Figure 6.4 Single, double, triple, and SuperJet grouting.

 Jet grouting 211

However, a considerable percentage of the grout is lost to the surface due to

the airlift. The double system has proven to be most effective in noncohe-
sive coarse-grained soils like gravel and sand.
The third method, which is called the triple system (T), utilises three
luids—grout, jetting water, and compressed air shrouding the water. This
system normally consists of a grouting nozzle below a water jetting nozzle,
added in order to convey as many excavated soil particles as possible to the
surface while limiting the grout ejected. While with the double system the
ratio between water and cement is ixed or can only be adjusted through
the water-cement content, the triple system achieves erosion and grout
injection independently and can thus be optimised for the required perfor-
mance. This has potential advantages in ine-grained and cohesive soils like
silt and clay, allowing adjustment of the cement content independent from
the energy used to cut the soil and where the double system would require
a separate pre-cutting step.
In the 1980s, experience and conidence with jet grouting spanned a
very wide range of application. Since the early 1990s, newer methods of jet
grouting capable of a considerably larger treatment range or column diam-
eter have been developed on grounds of cost and programme. This enabled
jet grouting to obtain a column with a diameter in excess of 5 m, or even
9 m in softer ground (Figures 6.5 and 6.6 show examples of such an over-
sized body). This method could improve volumes of soil 20 times as large as
the previous conventional systems, due to equipment development provid-
ing signiicantly higher low rates at higher pressures.
The successful construction of a large column requires the use of focused
jets, of which an example is shown in Figure 6.7, maintained in pristine
condition; otherwise a large proportion of the jetting energy is lost within
the system itself. Thus, jet grouting emerged capable of spanning a very
wide range of applications.
The results of jet grouting can vary according to both equipment and
soil types. Given these constraints, many measurements have been taken
by varying the values of key parameters as a basis of theoretical solutions;
however, even these trials cannot provide exact solutions because of the

Figure 6.5 Exposed SuperJet columns during technology development.

212 Ground improvement

Figure 6.6 Trial SuperJet columns.

limited investigation into the soil and a lack of understanding of the real
elementary process that occurs when the jet meets the soil.
In the late 1980s, a new concept provided an innovative progress for jet
grouting systems, namely, dual jets colliding with each other to limit their
eroding capability, thus achieving an exact intended diameter regardless of
soil type. The arrangement of these jets is shown in Figure 6.8a while an
exposed column is shown in Figure 6.8b.
The conceptual comparison of conventional and colliding methods is
shown on Figure 6.9, noncolliding jets producing columns of variable
diameter in variable ground. Colliding jet grouting has raised the required
design quality since its appearance under the name of ‘crossjet grout-
ing’. In the early 1990s, colliding jetting was further evolved to include
the deep mixing method to substantially increase the range of applica-
tion. Conventional in-situ soil mixing suffers from a serious drawback of
imperfect continuity when executed adjacent to walls; however, attaching
an assembly of colliding jetting equipment at the tip of a drilling bit or
blade as in Figure 6.10 has enabled the construction of optimal interlock-
ing, as shown in Figure 6.11.
Furthermore, the enhancement in this in-situ mixing system results in
more than four times the treated volume using the same equipment. This
is shown in Figure 6.12, the conceptual schematic of the jet and churning
system management (JACSMAN) system.

1m 2m

Figure 6.7 Focused jets.

 Jet grouting 213

(a)

Water jet
at 40 MPa
and 180 l/min

Cement-water slurry jet

at 4 MPa and 190–250 Eroded soil-water
l/min mixture
Cement-soil
mixture

(b)

Figure 6.8 (a) Dual colliding jets. (b) Columns produced by dual colliding jets.

6.3 THEORY OF JET GROUTING

Many factors inluence the eficiency and effectiveness of the jet grouting
process and require consideration when designing and constructing jet
grout columns.

6.3.1 Effect of dynamic pressure

When eroding soil with a high-pressure jet, the eroding distance radi-
cally increases after the pressure exceeds a certain level called the limit
break pressure (Figure 6.13). The erosion distance radically increases
after the jet pressure exceeds the unconined compressive strength of
both cohesive and sandy soil. Figure 6.14 illustrates the distribution of
dynamic pressure versus the distance from nozzle, in which the soil is
eroded to the distance at which the jet impact pressure attenuates to the
level equal to the unconined compressive strength of the soil (scaled
laboratory conditions).
214 Ground improvement

(a)

(b) (c)
Profile

N-value
Soil

20 40
10 30 50

Sand
and
gravel
Sand
and
clay

Fine
sand

Figure 6.9 Principles of cross jetting. (a) Strike marks of dual colliding jets on pressure-
sensitive ilm. (b) Soil cutting by a dual colliding jet. (c) Soil cutting by a single jet.
 Jet grouting 215

Figure 6.10 JACSMAN tool details.

Figure 6.11 JACSMAN column abutting sheet piling.

It is possible, with a lower impact pressure, to erode the same distance

over a longer time; however, the high pressure saves time for most practi-
cal applications. Typically, jet pressures between 30 and 60 MPa for an
overburden soil such as silt, sand, etc., and more than 200 MPa for rock
formation are employed.

Type A 3300 Type B 3700

1000 1300
2300

2300

650 100 500

1000

1800 1000 1400

Improving area Ap = 1.5 m2 Improving area Ap = 6.4 m2 Improving area Ap = 7.2 m2

Figure 6.12 Treated area enhanced by colliding jets in JACSMAN (all dimensions are in mm).
216 Ground improvement

Limit break
pressure (plimit)
Breaking speed ( vcut)

Minimum break
pressure (pmin)

Water jet pressure ( p)

Figure 6.13 Jet pressure regulated for soil erosion.

Sand Clay
Unconfined compressive Unconfined compressive
strength qu = 80 kN/m2 strength qu = 150 kN/m2
Eroding distance (mm)

Eroding distance (mm)

30 30
25 25
20 20
15 15
10 10
5 5
0 0
0 100 200 300 400 0 100 200 300 400
Jet pressure (kPa) Jet pressure (kPa)

Figure 6.14 Eroding distances vs. jet pressure for sandy and clayey soil (results of small-
scale laboratory tests).

6.3.2 Effect of flow rate

When pressurised luid passes through a circular nozzle, the following
equation is obtained from the law of conservation of energy:

p0 p
v0 = m 2 g = m 2 0 (6.1)
γ ρ
where p0: initial pressure at the nozzle, v0: initial velocity at the nozzle, g:
acceleration of gravity, m: nozzle eficiency, γ: luid unit weight, and W: luid
mass density.
 Jet grouting 217

Distance from the nozzle (cm)

1 5 10 50 100 300
1
Dynamic pressure ratio 0.5

0.1

0.05 Velocity of air

0m/sec
33m/sec
0.01 67m/sec
100m/sec
0.005
167m/sec
329m/sec
0.001

Figure 6.15 Dynamic pressure ratio (p/po) along jet centre axis with various air velocities.

A practical example of a calculation for pressure effect and low effect is

given here. If a water jet is discharged at 40 MPa through a ine nozzle of
2 mm in diameter such that the velocity of shrouded air is 100 m/sec, we can
obtain an eroding distance of 1 m (2 m in diameter) at the point of 4 MPa
from Figure 6.15 (dynamic pressure 0.1 times nozzle pressure po). This may
be regarded as the effective limit of the column for most practical purposes.
Since an excellent nozzle has m = 0.92 as an eficiency coeficient,
Equation 6.1 results in a low rate Q:
π 2
Q = vA = m 2 gp0 d (6.2)
4
= 49 l/min, where A: nozzle area, d: nozzle diameter.

If a 5-mm nozzle of the same eficiency is used instead, then in order to

achieve the same required column diameter, the low rate must be altered
in accordance with the square of the nozzle diameters:
2
Q1 d 
=  1
Q2  d2 
Hence the low rate is 306 l/min.

6.3.3 Effect of compressed air

An increased air velocity with even low pressure can extend the eroding
potential considerably, as illustrated in Figure 6.15 (dynamic pressure along
218 Ground improvement

jet centre axis with various air velocities). Jet grouting requires compressed
air for successful operation in several respects. It is irst indispensable for
obtaining maximum eroding energy and then of vital importance for con-
veying spoil up to the ground surface.

6.3.3.1 Effect of compressed air shrouding

A water jet as a ire extinguisher is totally effective; however, its effective-
ness is signiicantly decreased in water. Because jet grouting mostly treats
the soil beneath the groundwater level, a water jet alone cannot cause sig-
niicant ground improvement. In this respect, compressed air shrouding of
liquid jets is a primary technique in eliminating groundwater around the
jets, thus quasi-forming an atmospheric condition.
Figure 6.16 sketches the eroding distance of respective jets in air, in
water, and in water with an air shroud. This chart clearly demonstrates the
jetting principle that a liquid jet maintains a dynamic pressure ratio of 0.01
at a distance of 3 m in air. This distance is reduced to just 0.5 m in water;
however, with the addition of the compressed air around the water jet, it is
extended again to 1.1 ~ 1.2m.

6.3.3.2 The velocity and volume of compressed air

As stated previously, the mere presence of the air shroud does not always
prove successful, but it should also maintain a higher velocity than half

Distance from the nozzle (cm)

1 5 10 50 100 300
1

0.5

3 1
Dynamic pressure ratio

0.1

0.05
2

0.01

0.005 1 In the air

2 In the water with air jet
3 In the water
0.001

Figure 6.16 Dynamic pressure ratio (p/po) along jet centre axis.
 Jet grouting 219

the sonic velocity to ensure the formation of an atmospheric condition,

as is clearly outlined in Figure 6.15. Additionally, an air nozzle has to be
ring shaped or annular surrounding the nozzle, which preferably includes a
minimum straight length before the air discharge point.
The width of this annulus must be approximately 1mm thick as standard
which should provide suficient air low and yet does not allow any foreign
particles like sand to low upstream. Compressed air may be generated by a
low-pressure compressor rated at 0.7 MPa for work up to 20 m deep; how-
ever, a high-pressure compressor is required to withstand the groundwater
pressure for deeper works.

6.3.4 Effect of the soil

Soil type and stratigraphy inluence the quality of soilcrete (the soil-cement
product of jet grouting) and geometry of erosion. Figure 6.17 presents a
qualitative scale for soil ‘erodibilty’. In the local region of luid injection,
the turbulence created alone is enough to disaggregate cohesionless soil
types. As plasticity and stiffness increase, erodibility decreases to a point
where jet grouting may not effectively erode stiff cohesive soils.
Effective diameters of single luid jet grouting typically range from 300–
500 mm, and for double luid typically range from 800–1,300 mm.
The Japanese Jet Grouting Association proposes standard jet grout (soil-
crete) diameters for the triple luid (Table 6.1) and the SuperJet methods
(Table 6.2) to be used in different soil types.
In Table 6.2, the standard diameters listed represent special optimised
tooling. Conventional tooling may yield considerably smaller geometries. Soil
stratiication is also a consideration, as variable soil conditions lead to vari-
able soilcrete quality. Also, the jet grouting parameters may need to change
versus depth to create uniform geometry, or variable geometry may result.
Gravels, cobbles, and boulders, although considered cohesionless, may
range from highly erodible to very dificult to erode depending on in-situ
Low plasticity clays (stiff )
Low plasticity clays (soft)
Peats and organic silts

High plasticity clays

High plasticity silts
Dense clayey sands
Loose clayey sands
Low plasticity silts
Dense silty sands
Loose silty sands
Gravelly soils
Cobbly soils

Clean sands

Highly erodible Difficult to erode

Figure 6.17 Soil erodibility scale.

220 Ground improvement

Table 6.1 The standard soilcrete diameters (m) for sandy and clayey soil using triple
luid jet grouting
N valuea
Gravelb
Sandy soil N ≤ 30 30 < N 50 < N 100 < 150 < 175 <
≤ 50 ≤ 100 N ≤ 150 N ≤ 175 N ≤ 200
Clayey soil – N≤3 3< N ≤ 5 5< N ≤ 7 – 7<N≤9
Organic soilc
Effective diameter vs. depthd
0 < Z ≤ 30m 2.0 2.0 1.8 1.6 1.4 1.2
30 < Z ≤ 40m 1.8 1.8 1.6 1.4 1.2 1.0
Lifting rate (min/m) 16 20 20 25 25 25
Pumping rate (m /min)
3 0.18 0.18 0.18 0.14 0.14 0.14
Note: For cohesion c around 50 kN/m2, the standard diameter may be dificult to obtain. For sandy
soil with N > 150, and clayey soil with N > 7, the grouting speciications must be determined with
considerable examination such as a ield veriication trial.
a The largest SPT-N value of the soil to be treated should be used.
b For gravelly soils, the soilcrete diameter expected is tabulated diameter less 10%. A ield trial to
verify the diameter is recommended prior to production.
c For organic soil, considerable examination is recommended to determine the grouting speciications.
d For the depth Z > 40, considerable examination is recommended to determine the grouting
speciication.

density, soil matrix, and other conditions. When soil contains more than
30% gravel, or the size of gravels exceeds 10 cm in diameter (cobbles),
resulting soilcrete diameter may be smaller than expected. Reducing
tool rotation or increasing slurry pump rate may solve this type of prob-
lem. Boulders will block the jet stream and a ‘shadow’ of untreated soil
will exist beyond. Buried obstructions can also include trees, utilities, or
cemented soil.

6.3.5 Other effects

The quality of the material and internal inish of the nozzle is of vital impor-
tance as well as its dimensions and geometry. Furthermore, in reality, care
must be taken that even a perfect nozzle before use may be easily damaged
owing to anomalies in the jetting stream.
In order to account for this, inspection of the condition of nozzles
before and after each jet grouting operation has to take place. An optimal
inspection technique employs a special measurement system of dynamic
testing in association with pressure-sensitive ilms with a predetermined
range.
 Jet grouting 221

Table 6.2 The standard soilcrete diameters (m) for sandy and clayey soil using superjet
grouting
N value
Sandy soil N ≤ 50 50 < N ≤ 100 100 < N ≤ 150 150 < N
Clayey soil N≤3 3<N≤5 5<N≤7 7<N≤9
Organic soil
Effective diameter vs. depth
0 < Z ≤ 30m 5.0 4.5 4.0 3.5
30 < Z 4.5 4.0 3.5 3.0
Notes:
1. A ield trial prior to production works is recommended to verify the diameter.The tabulated diam-
eter is sometimes dificult to obtain in soil with certain characteristics.
2. For gravelly soil, a ield trial to verify the diameter must be performed prior to production works.
Soilcrete diameter with the tabulated diameter less 10% may be used for preliminary designing.
3. For soft soil (sandy soil N < 10, clayey soil N < 1), the diameter sometimes exceeds the tabulated
diameter, which leads to shortage of solidifying material and results in lower compressive strength
than expected. A ield trial to verify the diameter is recommended prior to production.
4. Sandy soil with N < 150 in the table applies only to nonsolidiied sandy soil. For clayey soil with
N > 9, you may consult the SuperJet Association for some speciications to achieve the construc-
tion objective. A ield trial to verify the diameter is recommended prior to production.
5. For cohesion c around 50 kN/m2, the standard diameter may be dificult to obtain.

If the jet is sound, the pressure-sensitive ilm reveals an annulus, with the
centre destroyed, which is the so-called core of the jet still maintaining suf-
icient eroding energy to penetrate the ilm, as sketched in Figure 6.18, left (a
focused low). For a defective jet, the ilm relects a totally coloured spot, with
no central penetration as sketched in Figure 6.18, right (a turbulent low).
Apart from dynamic pressure and low rate, there are other parameters
that have an inluence on the eroding power of a liquid jet. An experimental
equation explains this:

( )
−1 / 1.4
R = 4.95Kp0−1.4 Qw−1.6 N t−0. 2 vn−0.3 (6.3)

where:
R = Eroding distance (m)
K = Improvement factor (experience based) (m/sec)
p 0 = Pumping pressure (tonnes/m 2)
Qw = Flow rate (m3/sec)
Nt = Repetition frequency (number of times a jet nozzle passes the same
point)
vn = Rotational velocity of nozzle (m/sec) = Dm × π × Rs/60 (Dm: Diameter
of the monitor, R s: rpm of the monitor).

Figure 6.19 provides experimental results for the optimal repetition

frequency of the eroding jet, indicating that frequencies in excess of 5 only
222 Ground improvement

Figure 6.18 Focused low (left) and turbulent low (right).

4.0
Rs (Rotating speed: rpm)
3.5
Column radius (m)

3.0

2.5

2.0
Rs = 1.25
Rs = 2.5
1.5 Rs = 5
2.5
1.0
0 5 10 15 20
Repetition frequency

Figure 6.19 Experimental results for optimal repeating frequency of eroding jet.

marginally increase the column diameter. Lifting up the jetting rods in

steps provides the necessary rotation using an integral number, which is
not possible with a steady lift as shown in Figure 6.20 (lifting methods).
Each step corresponds to an intended diameter; however, practical experi-
ence gives a maximum 5 cm lift for up to 2 m in diameter, and a maximum
10 cm lift for more than 4 m in diameter, as optimal increments, but is
soil-type dependent.

6.3.6 Practical considerations

In order to successfully design a jet grout project, both theoretical and prac-
tical considerations need to be taken into account. For a successful project,
 Jet grouting 223

(a) (b)

t t

Figure 6.20 Lifting methods. (a) Intermittent lift. (b) Steady lift.

both columns must be installed correctly and the achieved properties must
be in accordance with those values required by the design.

6.3.6.1 Design parameters for jet grout material

Strength of treated ground is usually assessed on the basis of unconined
compressive strength tests on samples obtained by coring and/or in-situ
grab samples cast into moulds. The histograms shown in Figure 6.21 dem-
onstrate experiential unconined compressive strengths in granular (sandy)

1.3σ 1.3σ
50 50
x = 27.8 x = 127
40 40
Frequency

σ = 13.3 σ = 54.3
30 30
20 20
10 10
0 0
1 2 3 4 5 6 (MPa.) 20 60 120 140 180 220 260 (MPa.)
Unconfined compressive strength Unconfined compressive strength

Figure 6.21 Histograms of unconined compressive strength using triple luid jet grout-
ing in clayey soil (left) and sandy soil (right).
224 Ground improvement

and cohesive (clayey) soils. The Japan Jet Grouting Association has adopted
these distribution charts, deining the unconined compressive strength to
be taken for design to be the minimum safe values which range between
1% and 3% from the least values in the whole group. Table 6.3 shows the
standard properties for soilcrete using triple luid jet grouting.
This deinition gives the standard unconined compressive strengths as
follows (where the water/cement ratio of the grout is 1):

qu = 1 MN/m 2 (Unconined compressive strength in cohesive ground)

qu = 3 MN/m 2 (Unconined compressive strength in granular ground)

According to the German E DIN 4093 (draft) for the design of all kinds
of stabilised soil, the maximum allowable unconined compressive strength
(qu,k) to be used for jet grouting is:
qu,k < 10 MN/m2, which is calculated as the minimum of either the smallest
value measured in a series of four samples or (0.6 to 0.75) times the mean
value of the series of samples. The design value (qu,d) is: qu,d = qu,k × 0.85/γm
where qu,d is the design value, 0.85 is a factor to consider long-term load-
ing, and γm is the partial safety factor for this material. The partial factors
for the loads (load case 1) vary between 1.35 (for dead weight) and 1.5 (live
loads). The resulting global safety factor between mean value and design

Table 6.3 The standard properties for soilcrete using triple luid jet grouting
Unconined Modulus
compressive Bonding Tensile (E50) of
Grout strength Cohesion strength strength deformation
material Soil type (MN/m2) (MN/m2) (MN/m2) (MN/m2) (MN/m2)
JG-1(H) Sandy soil 3 0.5 1/3C 2/3C 300
Clayey soil 1 0.3 100
JG-1(L) Sandy soil 2 0.4 200
Clayey soil 0.7 0.2 70
JG-2 Sandy soil 3 0.5 300
JG-3 Sandy soil 1 0.3 100
JG-4 Organic soil 3 0.3 30
JG-5 Clayey soil 1 0.1
Notes:
1. All data are 28-day cured strength and were determined from core samples.
2. Strength-controlled soilcrete material is usually used for sandy soil. In case it is used for the soil
stratiied with sandy and clayey layers, the strength of clayey layers are reduced with the following
rates:
JG-2: 70% of JG-1
JG-3: 50% of JG-1
3. The densities of soilcrete are regarded to be similar to those of the in-situ soil.
4. For gravelly soil, sandy soil data are to be used.
5. Seven-day strength of soilcrete is regarded to be 30%–40% of the four-week strength.
 Jet grouting 225

strength from this is between 4.4 and 3.2. For this purpose, samples are
tested after a curing time of 28 days.
The directory from Japan Road Association states that permeability is in a
range of 1 × 10−6 to 1 × 10−7 cm/s. The uses of the grout material are as follows:

r JG-1: High-strength soilcrete (standard material)

r JG-2: Strength-controlled soilcrete (medium strength)
r JG-3: Strength-controlled soilcrete (low strength)
r JG-4: For organic soil
r JG-5: For clayey soil

The design standard strengths of cohesion, bond and tension in bending

are then determined with reference to the values shown in the Table 6.3. If
it is desired to use alternative values it is recommended that laboratory mix
design testing with representative soil precede production work.

PSI (MN/m2)
1500
(10.34)
Cement
content
1250 150-250
(8.61) kg/m3
Unconfined compressive strength

Sands and gravels

1000
(6.89)
200-275
kg/m3
750
(5.17)
Silts and silty sands
250-350
500 kg/m3
(3.44)
Clays
300-400
250
kg/m3
(1.72)
Organic silts and peats

0
0 10 20 30 40 50 60
Age (days)

Figure 6.22 Strength–material type chart. (Redrawn from Burke, G.K. (2004). Jet
grouting system: advantages and disadvantages, Proceedings of Sessions of the
GeoSupport Conference: Innovation and Cooperation in the Geo-Industry, Jan.
29–31, 2004, Orlando, Florida, United States; sponsored by International
Association of Foundation Drilling (ADSC) and the Geo-Institute of the
American Society of Civil Engineers.)
226 Ground improvement

Although very rare, soilcrete has been constructed to strengths in

excess of 20 MN/m 2 in clean sands using specially developed grout mixes.
An estimate of average strength for ordinary operations can be seen in
Figure 6.22.

6.3.6.2 Drilling tolerances

Drilling tolerances are particularly relevant with jet grouting as overlap-
ping of columns is vitally important. Inadequate interlocking not only takes
place through drilling deviation which increases the offset from a neigh-
bouring column with depth, but also through penetrating into a neighbour-
ing column that has already set. The latter problem results in jetting within
set and rigid material, consequently leading to unsuccessful works as no
column is formed, as diagrammed in Figure 6.23. Inadequate interlock-
ing can only be limited by excellent drilling coupled with in-hole survey
techniques. Because of this, jet-grouted holes should be surveyed whenever
possible to ensure deviation is within acceptable limits.
According to Japan Jet Grouting Association, drilled holes deeper than
30 m must have their inclines measured with an inclinometer or gyroscope.
If the incline exceeds the standard of 1:250, holes must be re-drilled. In
general, drilling tolerances of up to 1 in 100 can be achieved, but special
consideration for the speciic risks have to dictate the deinite tolerances for
the radius in the depth of interest.
–4.00
–3.50
–3.00
–2.50
–2.00
–1.50
–1.00
–0.50
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00

4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
y (m)

0.00
–0.50
–1.00
–1.50
–2.00
–2.50
–3.00
–3.50
x (m) –4.00

Figure 6.23 Effects and risks related to drilling deviations, and effect of poor drilling
tolerance on column construction.
 Jet grouting 227

6.3.6.3 Control of jet grout returns

The method of jet grouting uses hydraulic erosion to construct the column
geometry. The erosion media is either grout slurry or water, depending on
the system deployed. In order to control the in-situ erosion environment,
the borehole annulus must be an open pathway for return materials.
These return materials can vary greatly depending on the soils being
eroded and the erosion media. In some circumstances, it may be desirable
to ‘pre-cut’ or perform an erosion stroke with water only prior to jetting
with grout. This could be the case if

r A higher cement content (strength) was needed, feasible by eroding a

higher percentage of ine-grained soil from the desired depths.
r Assurance of protecting against surface heave, feasible if the returns
pathway is restricted.

Assurance of continuous returns during jet grouting is necessary for the

control of jet grouting and is a hallmark of quality jet grout construction.
Disregarding this requirement can result in nearby heave or settlement, poor
quality soilcrete, lack of geometry control, and impacts to nearby utilities.
Many things can be responsible for a loss of spoil return, such as the
following:

r A borehole restriction
r Too small a drill hole (annulus)
r Too small a hole through a footing
r Soft, squeezing clays
r Gravels that are collapsing
r Loss of air return
r Open, porous, gravely zones
r Fibrous peats
r Very soft clays
r Cohesive soil erosion
r Erodes in pieces that block the annulus
r Very thick (viscous) spoil

Spoil return can be enhanced by adjusting aspects of the jet grouting:

r Changing the grout viscosity

r Changing air pressure and low rate
r Use of casing to reduce up-hole friction
r Pre-cutting measures
r Auxiliary air-lift system
r Changing the borehole size
r Manual reaming of the borehole
r Reducing the jetting energy
228 Ground improvement

6.3.6.4 Sequence of construction

For every jet grouting project, there is a sequence of work that will provide
for differences in quality, deformations, returning spoils, and/or geometry.
The selection of installation sequence is important to the desired product
and, when working beneath or nearby structures, to the effect on them.
This sequence is selected based on experience and the most desirable
effect. The sequence for assuring continuity of a wall or base seal against
groundwater is different than what might be selected for the highest
strength. Similarly, when underpinning a structure or utility, the sequence
must be such that loads can be redistributed by arching to adjacent ground
until adequate strength can be developed by curing of the soilcrete.
Every case cannot be addressed here, but it is suficient to say that the
sequence is a planned approach that requires attention.

6.3.6.5 Quality control and validation

Section 6.3.6.3 alludes to problems of deviation but column diameter, posi-
tion, and properties must also be considered. Therefore, it is important
wherever possible to record and validate the installation of individual jet
grout columns. Most specialists have the instrumentation to record the fol-
lowing parameters during installation:

r Depth
r Withdrawal rate and
r Step height and step timing (considering rotation speed)
r Uniform lift rate (considering rotation speed)
r Air pressure and low rate
r Grout or water pressure and low rate
r Rotation speed
r Grout density

In addition, some specialists have developed inclinometers built into the jet
grout monitor that measure deviation of the drill string. It is also equally impor-
tant to carry out quality control testing on the grouts used. This normally
includes speciic gravity, viscosity, and strength by 28-day cube strengths.
The knowledge of all these parameters allows the site engineer to review
the column installation and come to a decision as to whether any column
is misplaced or incorrectly installed. This is of paramount importance for
base slabs or tunnel break-in or break-out where the omission or misplace-
ment of a column can have the most serious effect on performance or safety.
A further dificulty is the repair of these jet grout bodies as usually failures
are dificult to locate.
Franz (1972), Fritsch and Kirsch (2002), and Kirsch and Sondermann
(2002) list standards and publications relating to the control and execution of
 Jet grouting 229

jet grouting. Burke (2009) reviewed quality control considerations. Eurocode

EN 12716 is the European jet grouting standard code for execution.
Validation of jet grouting can be problematic. In order to fully validate
a project column diameter, position and strength or permeability must be
checked. Techniques typically carried out are as follows:
Column diameter: The most appropriate technique is to construct trial col-
umns and then expose them to measure diameter directly. This is
an excellent method but can only be used at shallow depths due
to the expense of accessing columns at depth. Coring of columns
can be successful but often suffers poor core recovery leading
to dificulty in interpretation of diameter or strength. Electronic
CPTs have been used to deine geometry as they can easily ‘feel’
the surface of a soilcrete column (Burke et al., 2003). In Japan
and Europe, thermocouples are being used to calculate diameter
by comparing energy from the binder hydration to the measured
in-situ energy (Meinhard et al., 2010). Borehole callipers can be
lowered and extended to measure the extent of a column prior to
initial set, but only in certain soil conditions by an experienced
contractor. Some geophysics companies are developing nonde-
structive techniques utilising 3D borehole radar, electric resistiv-
ity, and sonic response, as yet still remaining at the research stage
but offering a promising solution (Burke, 2012).
Column position: Column position relates to measurement of drilling toler-
ance, and as discussed above this is either accomplished by built-
in inclinometers or by survey of the hole prior to jetting.
Column properties: This is the most commonly measured using coring
techniques although some companies offer sampling within the
column prior to initial set. Some forms of nondestructive tech-
niques can be used as discussed above.
When working beneath or nearby structures and utilities, these items
should be monitored during all operations and alarms set such that the
operations are temporarily ceased if exceeded.

6.4 APPLICATION OF JET GROUTING

As set out in the introduction, jet grouting is an exceptionally versatile tool

when considering ground improvement as part of a project. There are many
applications that suit jet grouting but they can be grouped together as follows:

r Groundwater control
r Movement control
r Support
r Environmental
230 Ground improvement

Groundwater control applications include

r Preventing low either through the sides or into the base of an

excavation
r Controlling groundwater during tunnelling
r Preventing or reducing water seepage through a water retention struc-
ture such as a dam or lood defence structure
r Preventing or reducing contamination low through the ground

Movement control applications include

r Preventing ground or structure movement during excavation or

tunnelling
r Supporting the face or sides of a tunnel during construction or in the
long term
r Increasing the factor of safety of embankments or cuttings
r Providing support to piles or walls to prevent or reduce lateral
movement

Support applications include

r Underpinning buildings during excavation or tunnelling

r Improving the ground to prevent failure through inadequate bearing
r Transferring foundation load through weak material to a competent
strata

Environmental applications include

r Encapsulating contaminants in the ground to reduce or prevent con-

tamination off site or into sensitive water systems
r Providing lateral or vertical barriers to contaminant low
r Introducing reactive materials into the ground to treat speciic con-
taminants by creating permeable reactive barriers

These lists show that jet grouting has a multitude of uses, all of which
must be understood, designed, and executed accordingly. Some important
main applications are now described in more detail.

6.4.1 Groundwater control

The last three decades have seen an increasing number of large excavations
constructed in water-bearing soils. The use of conventional groundwater low-
ering techniques has been reduced as a result of the increasing importance of

r Economic water control

r Environmental aspects of the aquifer
 Jet grouting 231

r Observance of existing water rights

r Protection of existing buildings

Conventional chemical-based injection systems have been almost com-

pletely replaced by jet grouting techniques where the use of cement-based
grouts reduces alkalinity.
Typical waterprooing elements are vertical and horizontal walls with
and without an additional structural function in deep excavations, or for
dams and dikes, break-in and break-out blocks to assist tunnel-boring
machine operations. While with jet grouting columns a permeability of
10 −5 to 10 −6 m/sec can normally be expected, the permeability of the system
as a whole ranges from 10 −4 to 10 −5 m/sec. As a rule, the excavation cannot
commence until the allowable low rate has been achieved and proven by a
pumping test. Excess seepage is generally a result of a defect and can have
countless additional causes.
The detection and location of leaks is extremely dificult, sometimes
even impossible, and full or partial drawdown of the water table and the
observation of piezometers or the measurement of the ground temperature
during re-establishment of the water table are the most promising methods
of leak detection. The necessary remedial works are often time consuming
and extremely expensive, so the proper design and execution of jet grouting
sealing elements is vital to the success of the project. The design requires
the deinition of suficient strength and minimum permeability, homogene-
ity, and dimensional accuracy. It is essential to remember that water will
not forgive any mistakes.
Defects in jet grout bodies can occur as a result of

r Insuficient overlapping of individual jet grout columns

r Jet shadows caused by natural or man-made obstructions
r Inhomogeneities in the ground (hard layers embedded in sand, peat
layers)
r Instability and subsequent collapse of jet grout columns before they set
r Process deiciencies and interruptions, errors

To mitigate these risks, a thorough quality assurance plan is essential

and, indeed, is state of the art. The plan should include the following ele-
ments, also identiied in EN 12716:

r Setting out of the jet grout columns by x-y coordinates

r Drilling depth determined by eficient levelling systems
r Deinition of drilling and jetting parameters
r Execution of test columns, documentation and evaluation of results
r Deinition of the sequence of the works
r Identiication of obstacles and countermeasures
232 Ground improvement

r Grout composition and measurement of characteristics by sampling

at mixing station and in back low
r Measurement of drilling accuracy and countermeasures
r Process documentation during execution in real time of
r Speed of insertion and extraction of monitor
r Pressure and low rate of grout, water, and air
r Drilling and jetting rotation
r Data secured on memory cards and modem transferred to backup
systems

When looking at the evolution of grouting techniques in contractual

terms, it is clear how much injection of sediments has departed from rock
grouting. It must be remembered that the completed jet grout body is not
homogeneous and therefore generally does not exhibit a constant strength
or hydraulic characteristic. Design, speciications, and quality control must
therefore relect an uneven distribution of strength and permeability due to
the variability of the soil under treatment.
Horizontal jet grout barriers in deep excavations should therefore be
designed and executed with the following considerations:

r Minimum slab thickness not to be less than 1.0 m and to be increased

by 0.1 m for every metre in excess of 10 m depth for safe uplift slabs
r Large slab areas to be divided into compartments of 2000 m 2
r Increase of slab thickness in the immediate vicinity of vertical walls
r Avoid different slab elevations in one compartment
r Avoid location of slab within unsuitable soil conditions
r Time schedule to allow for possible remedial work
r Avoid anchored jet grout slabs
r Prepare emergency plan

Similar recommendations apply for vertical jet grout barriers as structural

members:

r Applications with water pressures in excess of 5 m require special

attention (redundant design, appropriate checking procedures, emer-
gency plan)
r Identify soils with erosion potential in case of leakages
r Avoid slender construction elements
r Special care required when ground anchors are necessary

6.4.2 Underpinning
Underpinning of structures using jet grout normally involves the construc-
tion of a body of improved ground beneath the structure such that the
 Jet grouting 233

+26.10 mWN
±0.00 m

Silt and sand

−3.00 m
First cut −3.50 m
−4.00 m
L = 14.7
0m
Alpha =
10°
Second cut −7.00 m 208.16 kN/m
−7.50 m L = 17.60 m
Alpha = 5°
416.59 kN/m
−9.78 m
Soilcrete

−11.78 m
260

Figure 6.24 Jet grout underpinning adjacent to an excavation in London.

structural load is transferred to depth. If the underpinning is carried out

next to an excavation, then the jet grout body must be designed accord-
ingly and the stability checked for bearing capacity, sliding, and overturn-
ing. There is sometimes an economic relationship between the creations
of a gravity underpin (i.e., a body that is self-supporting and stable) and
a propped or strutted body where overturning or sliding is restrained by
props or anchors (as for the case history below).
An example of jet grout underpinning adjacent to an excavation is shown in
Figures 6.24 and 6.25. This example is taken from a project in London where a
self-supporting underpin was required adjacent to a new basement construction.
The design of a jet-grouted underpin is exactly similar as for any gravity struc-
ture except that consideration needs to be taken into account that the strength
of the jet grout body is usually signiicantly lower than brick or concrete.

6.4.3 Tunnels and shafts

6.4.3.1 Bottom slab
Base sealing of the slabs of shafts for tunnelling can be designed for the
application of jet grouting to prevent base heave or piping in cohesionless
soils saturated with groundwater. As discussed in Section 6.4.1, these con-
structions are risky if incorrectly executed and require careful design and
application.
234 Ground improvement

Figure 6.25 Exposed jet grout underpinning adjacent to an excavation in London.

The Academia of Japan dictates that normally the thickness of this slab
must exceed half of the span between shaft walls. This is not the case out-
side of Japan, where sealing slabs have incorporated tie-down anchors, or
were placed well below the excavation depth, to counter the buoyant forces.
Thinner slabs are possible by employing circular arc beams on which only
compressive stress acts, as illustrated in Figure 6.26. This method of design
results in an arch prop 3 m in thickness even at a position of 40 m below
ground level.
A tentative calculation gives a maximum value of 1.1 MN/m 2 and a mini-
mal value of 0.95 MN/m 2 as compressive stress on both sides of the arch. As
the average unconined compressive strength of treated soils by jet grouting
commonly exceeds 3 MN/m 2 , this gives a high assurance of success.

6.4.3.2 Subsurface props

Displacement of walls is always of primary concern in open excavation.
Late propping during excavation often causes tilting and/or settlement of
not only adjacent buildings but water supply, sewer lines and other under-
ground facilities. Therefore, jet grouting offers the radically different
approach of an in-situ soil-mix propping prior to excavation.
A practical case history briely explains the result. The work required an
excavation of 10 m depth in a soft clayey layer for basement construction,
but adjacent houses were so close to the site that they were afraid of being
largely undermined due to displacement of walls for shoring, as shown in
Figure 6.27. Consequently, jet grouting-produced props of just 1-m thick-
ness at the bottom of excavation have proved successful together with a
row of conventional strutting at ground level. Adding a row of grouted
 Jet grouting 235

GL

WL GL

WL

40.0
–40.0
VA VA 1.5 1.5 3.0
NA NB

All dimensions in m
12.0
(a) (b)

Figure 6.26 Base sealing of a shaft. (a) Conceptual cross-section. (b) Layout of jet-
grouted arch.

props enabled the reduction of displacement by approximately 80%, as

clearly shown in Figure 6.27.

6.4.3.3 Roof barriers

In starting a tunnel-boring machine (TBM) through a wall of a shaft into
an alluvial deposit, the soil surrounding the TBM may be lower in strength
due to the loosening effect of the construction of the structure. This could
trigger collapse or settlement because of extension of this loosening to the
ground surface, especially in the case of shallow tunnelling. Given such
dificulties, jet grouting offers theoretical advantages in designing roof bar-
riers. The design geometry is explained by reference to Figure 6.28, which
illustrates how to obtain the zone to be treated (R–a), the property of which
is to be reinforced by jet grouting.
A successful design follows an achieved line of shear strength to exceed a fail-
ure envelope of Mohr circle of the original ground. Figure 6.28 also shows that
the radial and tangential stresses balance each other on the boundary line of
the elastic region from the plastic one, and consequently derives Equation 6.4:

∂σ r σ − σr
= θ (6.4)
∂r r
where Xr = radial stress, Xϕ = tangential stress, and r = variable radius.
236 Ground improvement

Displacement (mm)
Soil SPT
OUT IN
0 0 50 –20 0 70
Short strut
GL–3.0m
Actual
Calculation (with grouted props)
(with grouted
props) NS
Depth (m)

GL–9.1m
10
Grouted props (1 m thick)
RN
Calculation (with no grout prop)
SN
SS
NS
Soils description
20
NS –Clayey silt
SN –Silt Clay
RN –Clayey silt with gravel
SS –Fine sand

Figure 6.27 Comparison of calculated and actual wall displacements.

Next, since a failure takes place when the failure envelope becomes hori-
zontal and the internal friction angle becomes zero, Equation 6.5 is derived as:

σ θ − σ r = 2c (6.5)
where c = cohesion.

σθ σr

H
R

a Π
+φ/2
ϕ
φ

Figure 6.28 Roof barrier of a tunnel.

 Jet grouting 237

Then, substituting boundary conditions into the above simultaneous

equations to obtain the plastic region leads to the following equation:

 R γ
(
ln   = t H − R
 a 2c
) (6.6)

where R = plastic region, Lt = average unit weight of the soil, H = depth to

the centre of the tunnel, and a = radius of the tunnel.

6.4.4 Environmental applications

One of the more interesting uses of jet grouting is in the environmental
ield. There are many applications based on the ability of jet grouting to
form bodies at considerable depth while only requiring small penetrating
drill holes. The main uses can be classiied as follows.

6.4.4.1 Encapsulation
Achieving encapsulation of contaminants at depths where conventional
excavation would be dificult, as for the example shown in Figure 6.29.
Additionally, the grouted body is usually more impermeable than with con-
ventionally grouted ground, leading to more security in contaminant control.

Trench
backfill

Brick
culvert

Historic
grout sealing

Location of
phenolic
contamination

Jet grout
columns

Figure 6.29 Encapsulation of contaminants at depth.

238 Ground improvement

6.4.4.2 Containment barriers

In dificult ground, jet grouting can provide an effective method of creat-
ing a barrier as was achieved in a sand seam at a landill site in Dundee,
Michigan, United States in 2005 (Burke, 2007). Its main advantage is the
ability to be selective in which zone is to be cut off. This is especially advan-
tageous for deep applications. The grout or water jet (depending on system
utilised) scours weak and loose material, penetrating into issures and frac-
tures and replacing the permeable inill material with relatively imperme-
able grout. The effective distance penetrated will depend on the system and
rock type but has been shown on a number of projects to be effective up to
1 m from the hole position. Lateral barriers are typically speciied in terms
of permeability, and for rock it is usually possible to achieve 5–10 times
lower permeability than using conventional rock grouting. For soils, the
reduction when compared to permeation grouting can be as high as 10–50
times. As with all jet grout projects consisting of barriers or bodies con-
structed from interlocking columns, care must be taken during construc-
tion to minimise deviation from design locations, and this should always be
taken into account when designing the scheme.
At the Dundee, Michigan project, an industrial manufacturer needed a
groundwater containment barrier on the down gradient side of a disposal
area (Figure 6.30). A 360-m-long wall was needed, with a requirement for
thickness = 0.9 m, strength = 1,034 kPa minimum, and permeability < 1 ×
10 −6 cm/sec. Although the focus was on a sand seam at a depth of 6–9 m,
the treatment zone was speciied to treat signiicantly above and below the
sand seam to ensure full treatment and cutoff. Site access and mobility
along the wall was severely restricted, preventing most conventional meth-
ods of wall construction.

221 m MW–42

213 m
MW–341
205 m
Building

MW–321
River
MW–47

MW–52
Jet grout
MW–48 Barrier location

Figure 6.30 Plan view of jet-grouted soilcrete barrier wall.

 Jet grouting 239

Figure 6.31 Dual axis jet grouting operation for the jet-grouted soilcrete barrier wall.

The double luid system of jet grouting was used, utilising a dual axis rig
(Figure 6.31). This enabled two columns to be constructed concurrently
from the horsepower of a single grout pump. Pairs of columns were con-
structed ‘fresh-in-fresh’, meaning that set was intentionally not required so
to ensure jetting energy connected with adjacent work to preclude leaving
windows in the wall. This raised the certainty of closure of the barrier and
left only drilling verticality as a potential problem for the continuous wall.
Slowed drilling penetration and high speed rotation ensured vertical drill-
ing after setup.

6.4.4.3 Active barriers

In recent years, jet grouting has been used in the construction of permeable
reactive barriers (PRB). These barriers contain materials that react with spe-
ciic contaminants such that they are rendered harmless or less dangerous.
Typical materials are zero valent iron (ZVI), granulated active carbon
(GAC), or biologically active (BA). Design of these barriers is beyond
the scope of this chapter. A guide was published by the United Kingdom
Environmental Agency in 2002. To construct these barriers, the reactive
material is either introduced in place of the grout, or a cavity is created
by jetting and is tremie illed with a prepared material, as was the case in
Memphis, Tennessee, United States in 2006. The process is illustrated in
Figure 6.32 and described below.
ZVI was installed in a PRB to reduce levels of chlorinated volatile
organic compounds (CVOCs) in a deep luvial aquifer at the former Defense
Distribution Depot known as the Memphis Depot (Endo, 2009). The bot-
tom luvial aquifer is approximately 24.3 m below ground surface and is
1.8–2.4 m thick at the PRB location. The luvial aquifer is composed of
sand, sandy gravel, and gravelly sand. Groundwater in the luvial aquifer
 240
Ground improvement
Schematic of ZVI PRB installation
(1) Advance boring to the (2) Begin soil erosion/column (3) Complete column; use revertable (4) Using treme, displace guar
top of clay development slurry to maintain geometry slurry with reactive media
Batch mix iron and
sand mix
Waste
containment
vessel

Treme pipe

Figure 6.32 ZVI-PRB installation processes.

 Jet grouting 241

beneath the Memphis Depot has historically contained 1, 1, 2, 2-tetra-

chloroethane (1, 1, 2, 2-PCA) up to 41,000 micrograms per liter (μg/l) and
trichloroethane (TCE) up to 7,100 g/l. The PRB at Memphis Depot was
expected to be capable of treating the CVOCs to drinking water standards.
The jet grouting at the Memphis Depot was performed in two phases.
Phase I involved the construction of the geometry of the PRB using
conventional jet grouting methodologies with a water and a revertible
(biodegradable guar drilling luid), which was used to erode and remove
the soils. Phase II involved mixing the iron with sand and placing it down
the hole via a tremie pipe; the enzyme required for breaking the guar
slurry was also added during this phase. Because of its higher speciic
gravity, the sand and iron mix displaced the guar/water/soil mix within
the jetted geometry.

6.4.5 Waterfront structures

With aging waterfront structures and the need for deep water berthing, jet
grouting has emerged as an economic solution. Jet grouting is a system that
can work around the many buried features (sheetpiling, anchors, deadmen,
piling) to reduce the loads on existing walls and improve the stability of the
in-situ system (Figure 6.33). It can target speciic locations and depths to
overlap and offer vertical and horizontal support.

Gantry crane track

10
Anchor

Sheet pile Sheetpile

–10

Abandoned Current
sheet pile –20
Piles mud
line
–30

–40
New
Soilcrete dredge
line
–50

Figure 6.33 Typical port cross-section and solution for berth deepening.
242 Ground improvement

6.5 CASE HISTORIES

6.5.1 Seismic remediation: Wickiup Dam,

La Pine, Oregon, United States
SuperJet grouting provided seismic remediation of liqueiable soils layers
within a dam embankment in western Oregon. Wickiup Dam is a zoned,
rolled, earthill embankment with a main river embankment section height
of 30.48 m and a crest elevation of 1,324.97 m. The main embankment
transitions into a 4.83-km-long, 12.19-m-high dike section on the left
abutment. The dam is founded on bedrock while the left abutment dike is
founded on deep and bedded luviolacustrine deposits.
Analysis of the foundation materials in the left abutment dike indicated
that two separate layers of diatomaceous silt and one layer of volcanic ash
are likely to liquefy if the dam is subjected to the design earthquake. Superjet
grouting allowed the liqueiable foundation soils to be targeted for stabilisa-
tion in situ, allowing normal reservoir operations and reducing the inherent
risks associated with an excavate-and-replace alternative (Figures 6.34 and
6.35). Additionally, the jet grouting programme signiicantly reduced the
environmental impacts of construction operations on the pristine location.

Figure 6.34 SuperJet grouting in progress at Wickiup Dam.

 Jet grouting 243

Downstream berm
Existing dam placed after jet grouting
Existing ground
0 m (ft)
Gravel with sand 3.9 m
(13 ft) 8.2 (27)
Sand 12.1 (40)
Volcanic ash
15.2 (50)
Diatomaceous silt
Dense silt and sand 18.2 (60)
19.8 (65)
Diatomaceous silt Soilcrete
26.2 (86)
Dense silt and sand

Figure 6.35 Section view of left abutment dike at Wickiup Dam.

A total of 854 soilcrete columns were constructed along a 685.80 m length

of the dam toe. The columns were spaced on 3.96 m centres and each had a
diameter of 4.27 m. The treatment was performed within the depth interval of
13.41–26.52 m at the northwest end of the alignment, and sloped upward to
the depth interval of 2.44–6.10 m at the southeast end. Over 68,152,238 l of
grout slurry was pumped, and the volume of the soilcrete exceeded 153,675 m3.
A preproduction test programme veriied soilcrete geometry and qual-
ity prior to inal design and construction. The test SuperJet columns were
2500
(17.23)
Volcanic ash
Upper diatomaceous silt
Unconfined compressive strength PSI (MN/m2)

Lower diatomaceous silt

2000
(13.78) Dense silt and sand

1500
(10.34)

AVG
AVG
1000
(6.89)
AVG AVG

500
(3.44)

Target 28 day strength = 200 psi

(1.37 MN/m2)
0
Foundation material type

Figure 6.36 Unconined compressive strength test results from Wickiup Dam. The
owner required a minimum speciic gravity, which kept the strength higher
than what the speciication called for.
244 Ground improvement

constructed in an array with centre-to-centre spacing ranging from 3.51–

3.96 m to determine optimum achievable column diameter in the target soils.
Computerised data collection of all jet grouting parameters during
installation veriied that project procedures were met. An excavated trough
directed spoil from the drill hole to nearby pits. These spoils solidiied and
were excavated and used as ill material for a planned downstream berm.
Core holes drilled at strategic locations within the array determined the phys-
ical characteristics of the soilcrete, such as strength, fracture density, air vesicle
and soil inclusion volumes. Laboratory tests of the core samples conirmed that
soilcrete strengths exceeded the target 28-day qu of 1.37 MN/m2 (Figure 6.36).

6.5.2 Compression ring access shaft: Grand & Bates

Sewer Relief, St. Louis, Missouri, United States
Triple luid jet grouting was used to construct two access shafts in clayey
soils for the Metropolitan St. Louis Sewer District’s Grand & Bates Sewer
Relief project (Figure 6.37) (Camper, 2002). The close proximity to road-
ways and homes was a primary concern.
The shafts were composed of interlocking soilcrete columns forming a
continuous soilcrete compression ring (Figure 6.38). Shaft A had an inter-
nal diameter of 10.97 m and a depth of 20.88 m. Shaft B had an internal
diameter of 9.75 m and a depth of 16.46 m, using full- and half-circle col-
umns. Nominal column diameters of 1.07 m were achieved.
The 1.07-m-diameter columns were on a 0.91-m-centre spacing along the
circumference of the access shaft. A primary and secondary sequence of
installation was used (no adjacent columns installed on the same day) to pro-
vide for the highest quality soilcrete. Half columns with 1.07-m to 1.22-m
nominal diameters were strategically installed behind the interstice area of
the full columns to provide the required wall thickness and help assure effec-
tive water cutoff. An inner shotcrete facing was applied as excavation of the
completed shafts progressed (Figure 6.39). The shotcrete facing protected
the soilcrete columns from the elements and retained any loosened soilcrete.
Both shafts received rock grouting and shear reinforcement at 31 loca-
tions. Core holes were drilled through the soilcrete columns and 6.10 m into
the rock. The cores conirmed soilcrete quality and provided a drill hole for
a pinned connection between the soilcrete columns and the rock. The core
holes were water tested, and then pressure grouted, if necessary. Pressure
grouting created a grout curtain in the fractured rock below the soilcrete
columns. A 9.14-m-long #8 reinforcing bar was then tremie grouted into
each column. The bar extended 3.05 m up from the bottom of each column,
and 6.10 m into the rock (Figure 6.40).
Two full test columns and one half-test column were installed at production
column locations prior to the start of production work at each shaft location.
These tests helped to conirm the jet grouting parameters necessary to create the
 Jet grouting 245

Jet grouted soilcrete

access shaft Jet grouting soilcrete
Drop shafts construction shaft
T8D T6D Outlet shaft
T4D T3D T2D and structure
T11D T9D T7D T5D
T10D

Overburden soil
(Loessial alluvial deposits)
T1D
Original planned tunnel location
LIMESTONE
BEDROCK
As-built tunnel location

Figure 6.37 Section view of tunnel alignment with soilcrete access shaft locations.

N
Legend

Full soilcrete columns (41 EA)

3 ft (0.9 m) soilcrete
Half soilcrete columns (41 EA)
column diameter
(TVP) 3 ft (0.9 m) CTC Core locations (4 EA)
spacing (TVP) Anchor keys (28 EA)
36 ft (10.9 m) I.D.

Shotcrete over wire

mesh for protection

Figure 6.38 Plan view of soilcrete access shaft.

Figure 6.39 Shaft excavation in progress.

246 Ground improvement

Ground water
level 5.5 m (18 ft) Minimum 5 cm
(2 in) shotcrete
Soilcrete and welded wire Embedded

20.8 m (68.5 ft)

half column mesh lining strain gauges

Soilcrete
full column
Loessial & alluvial
deposits
Bottom of soilcrete
34.4 m (113 ft)
9.1 m (30 ft)

Fractured #8 bar shear key

rock

#8 × 6'-10''
rock bolts at
1.5 m (5 ft)
Limestone 9.7 m (32 ft) ID
c-c vertical
bedrock
spacing

Bottom of shaft

Figure 6.40 Section view of soilcrete access shaft.

required geometry. The test columns were installed using early setting cement
so that veriication coring could be performed three days after installation.
During production, neat cement grout and in-situ soilcrete samples were
taken once daily and cast for UCS testing. Average compression test results
for both shafts were 6.13 MN/m2 for wet soilcrete samples at 28 days. Once
the columns achieved adequate strength, four continuous 85-mm-diameter
soilcrete core samples were taken from columns equidistant around the shaft
circumference. The irst core hole was taken at the interstice area of the test
columns. The three remaining core holes were taken from production col-
umns at their interstice areas. A total of 32 core samples from each shaft were
tested. Eight samples were tested at 28 days for each core-hole location. The
average UCS test result of the cores from each shaft was 5.86 MN/m2 at 28
days, offering good correlation with the wet samples. All core-hole locations
were illed with neat cement grout after this testing was performed.
Vertical ground movements of the surrounding area were monitored
before and after each production shift. No ground heave was detected.
Twelve vibrating wire strain gauges were installed into each full column at
both shafts immediately after grouting. The strain gauges were installed at
three elevations for a comprehensive reading (Figure 6.41).
 150/1 0
130/.89 Measured hoop stress (psi/mpa, 55 ft/16m level)
10/3.0
110/.75 Excavation depth
20/6.0
positive hoop stress indicates compression;

90/.62
70/.48 30/9.1
Measured hoop stress psi/mpa

negative indicates tension

Excavation depth (ft/m)

50/.34 40/12.1
30/.20 50/15.0
10/.06
60/18.0
–10/.06
–30/.20 70/21.0

–50/.34 80/24.3
–70/.48 90/27.4
–90/.62
100/30.4
–110/.75
–130/.89 110/33.5

–150/1 120/36.5
7/19
7/29
8/5
8/13
8/21
8/29
9/11
9/19
9/27
10/7
10/15
10/23
10/31
11/8
11/19
11/27
12/4
12/12
12/19/2002
12/26/2002
1/5/2003
1/13/2003
1/21/2003
1/29/2003
6-Feb-03
14-Feb-03
24-Feb-03
2-Mar-03

Jet grouting
Date
Figure 6.41 Measured hoop stress provided from strain gauges.

247
248 Ground improvement

6.5.3 Groundwater control: Waste-water treatment

plant, Providence, Rhode Island, United States
A circular jet-grouted cofferdam access shaft (Figure 6.42) was con-
structed as part of the rehabilitation of a 75-year-old primary water sup-
ply conduit for the city of Providence (Oakland et al., 2002). The shaft
was needed over the cast-in-place portion to access the entire system for
inspection and repairs. The shaft did not require bracing, which reduced
the construction time substantially and allowed an unobstructed shaft
for access to the tunnel. The soilcrete wall formed a stable, erosion-
resistant surface capable of being disinfected quickly in the event the
tunnel unexpectedly had to be put back into service. The conduit could
only be accessed within the footprint of the city’s settling basin. After
repairs to the conduit, the shaft served as the excavation for a planned
valve structure.
The targeted soils consisted of 5.2 m of very dense, gray ine to coarse
sand with various amounts of silt and gravel underlain by granite bedrock
that was slightly weathered at the top with a 0.3 m sand seam approxi-
mately 0.3 m below the top of the rock. Jet grout holes were predrilled,
and the jet-grouted cofferdam columns were doweled into the top of the
bedrock to ensure the seal (Figures 6.43 and 6.44).

5.2 m

Existing 2.2 m 2.2 m 3.2 m 3.6 m

aqueduct 2.7 m

Columns to top of pipe

Half columns to cut

Ø 3.0 m off flow below pipe
Ø 1.5 m

Figure 6.42 Plan view of jet-grouted soilcrete column layout.

 Jet grouting 249

Soilcrete
Unconfined compressive
strength = 3.44 MN/m2

Granite

Figure 6.43 Cross-section of jet-grouted soilcrete cofferdam around and below the
existing conduit.

Figure 6.44 Soilcrete seal around and below the conduit with the bulkhead visible within
the conduit.
250 Ground improvement

6.5.4 Underpinning: Vassar College,

Poughkeepsie, New York, United States
Avery Hall, a 140-year-old historic building at Vassar College, was sched-
uled for demolition and replacement, with the exception of the West
Façade, which was to be incorporated into a new building (Burke, 2007).
A new mat foundation, requiring excavation, was to be constructed for
the replacement building. Jet grouting combined with temporary tieback
anchors was chosen for both the underpinning of the West Façade and for
temporary excavation support (Figures 6.45 and 6.46).
By constructing overlapping soilcrete columns, the footprint of the exist-
ing foundation was extended to within 0.3 m of the top of the clay layer
located 6 m below grade. The soilcrete underpinning wall provided support

Figure 6.45 Excavated Avery Hall façade underpinned with jet-grouted soilcrete
columns.

(a) (b)
28”
Jet grouted soilcrete 13”
underpinning column Working (exist) grade
(typ)
Tieback

Soilcrete

Bottom of excavation

Figure 6.46 (a) Partial plan view of jet-grouted soilcrete underpinning columns beneath
Avery Hall façade. (b) Section view.
 Jet grouting 251

by bearing below the excavation of the new mat foundation, and provided
the necessary temporary excavation support along the façade.
The grouting extended to a minimum of 0.6 m above the bottom of the
foundation in order to bind the rubble foundation together. Work was
sequenced such that any movements would be minimised. Before produc-
tion work, a test programme consisting of the installation of two soilcrete
columns was performed to enable the grouting contractor to conirm or
adjust the jet grouting parameters to ensure column geometry.

6.5.5 Environmental: Philadelphia Airport,

Runway 8-26, Pennsylvania, United States
A new commuter runway (Runway 8-26), was planned for construction at
the Philadelphia International Airport. A 300-m portion of the runway was
to be constructed over the Enterprise Avenue Landill (Furth et al., 1997).
The landill mass consists of 6 m of incinerator ash and concrete, asphalt,
rock, metal, and hazardous materials. The site was clay capped in the 1980s,
but additional remediation was needed to meet US Environmental Protection
Agency inal closure requirements before the runway could be constructed.
One component of the closure plan included installation of a low-perme-
ability horizontal barrier above a very thin (approximately 0.61–0.91 m)
natural clay stratum which underlies an approximately 1,020 m2 area of the
landill footprint so as to ensure that a minimum 1.52-m-thick low-perme-
ability barrier exists beneath the entire 150,000 m2 landill (Figure 6.47).
The new barrier was constructed using double luid jet grouting to achieve
remote excavation and replacement of the bottom 0.91 m of the waste mass
with a low-permeability soilcrete. The grout slurry was formulated to meet

2-7m future
embankment fill

Clay cap (1.5 m)

Incinerator ash and Interconnected

landfill materials (6 m) soilcrete columns
(1.67m O.C.)

Clay stratum Min.

1.5m

Figure 6.47 Jet-grouted soilcrete column proile view.

252 Ground improvement

the low-permeability (1 × 10 −9 m/sec), low-elastic modulus (124,100 kPa)

and compressive strength requirements (900 kPa) for the project design.

6.5.6 Settlement control: Japan

A tunnel was required to be constructed beneath a street under which were
buried numerous services. In addition, the adjacent buildings were sensi-
tive to movement. The solution adopted was to construct a heading from
spiles (horizontal piles) supported on jet grout columns toed into competent
ground. In this way the jet grouting supported the tunnel drive and reduced
settlements to acceptable levels. The small diameter holes required to install
the columns were also of beneit in penetrating between the services. The
crossjet system was chosen as the ground conditions were variable and with
this system the column diameter could be guaranteed (Figure 6.48).
In this project, an arrival shaft of a shield tunnelling machine was sup-
ported with many anchors, which had the potential to hinder the machine’s
arrival at the shaft, and therefore needed to be removed. At part of the wall
where the shield reaches the shaft, it was decided to remove the anchors
prior to the arrival of the machine and improve and stabilise the soil behind
the shaft wall. Figure 6.49 shows the exposed soilcrete at the opened wall.

6.5.7 Cofferdam sealing: New Orleans,

Louisiana, United States
At two locations where levee walls were breached during Hurricane
Katrina, the repair required that new sheetpile walls be constructed to

Building Building

Buried pipe Buried pipe

Fore piles
6000

Supporting Crossjet grouting

Primary executed columns

Gravel
7000
7854

Clay

Sand

13136

Figure 6.48 Settlement control for tunnelling, Japan.

 Jet grouting 253

Figure 6.49 The exposed soilcrete at the opened wall.

enable removal and replacement of the emergency ill (Burke, 2007). In

each location several groups of jet grout columns were constructed to con-
nect and seal the space between the old and the new sheetpiling. In each
location, the double luid system of jet grouting was used to create the soil-
crete columns, which hardened to provide an excellent groundwater seal
(Figure 6.50).
Jet grouting was employed for its ability to easily access these locations
on hastily constructed soil berms, and the ability to cleanly erode the soft
organic soils and encapsulate the sheetpile sections with a high-strength
low-permeability soilcrete.

6.5.8 Waterfront structure: Battery Park City

Authority, New York City, United States
Battery Park City Authority, on the Hudson River, is a combined residential/
commercial development built on land ‘created’ from material excavated
during the construction of the World Trade Center (Boehm, 2004). Further
development in the 1970s included the construction of a 21.3-m-wide river-
front esplanade consisting of a reinforced concrete relieving platform sup-
porting sand ill. Parallel to the river, the esplanade supports vertical timber
sheeting to retain up to 1.8 m of soil. Recent improvements in Hudson
River water quality resulted in an increase in the Teredo Navalis mollusk
population. These worm-like, marine borers were attacking and destroying
the timber sheeting.
Because borer activity would eventually result in loss of soil and surface
subsidence, replacing or supplementing the timber sheeting was imperative.
However, extensive development of the area, limited workspace, and dif-
icult subsurface conditions precluded conventional construction methods.
 254
Ground improvement
Minimum breakback of Land side Minimum breakback of
2.1 m from new 1.0m 0.6m 0.6m 1.0m 2.1 m from new
sheet pile tie in to 2.3m sheet pile tie in to
2.3m

breach near mirabeau ave

allow drill placement 0.7m 0.7m allow drill placement
for jet grouting 0.3m 0.3m 0.3m for jet grouting
MA54
0.3m MA58 MA58
MA54
Existing I wall Existing I wall
MA54 MA58
MA54 1.1m 1.1m MA54
0.5m 0.6m MA54 MA58
0.6m 0.5m
MA54 MA58

0.3m 0.3m
0.4m 0.4m
0.8m 0.8m
PZ-27 sheet pile
0.5m (repair wall) 0.5m

Canal side
Figure 6.50 Plan view of jet-grouted soilcrete locations for sheetpile seal.
 Existing grade
north meadow
Wall (varies)
Sand fill
El. –0.7 m
MHW Concrete platform
Reinforced concrete pile cap Concrete beam Minimum limit of
El. –0.6 m continuous grout wall
MLW Varies
45.7 cm Timber sheeting
rip-rap
(typ)
–2.1 m Sand

Fill (typ)

50.8 m square El –4.1 m Coarse stone

concrete piles (typ)

Jet grouting
Figure 6.51 Proile of jet-grouted soilcrete wall at Battery Park City Authority.

255
256 Ground improvement

Jet grouting provided an effective alternative, since jet grouting can be

readily accomplished in conined spaces and is effective across the widest
range of soil types.
The jet grouting work was completed in two phases. While the irst
phase work area was relatively open, the second phase was located within
extremely restrictive, urban surroundings, requiring special attention to
site conditions and spoil containment and disposal.
Project requirements on each phase called for supplementing the timber
sheeting with an in-situ, jet-grouted structural wall, placed directly behind
and in contact with the timber sheeting. The subsurface proile consists of
sand backill placed over ilter stone. This in turn is underlain by a layer
of crusher-run quarry stone containing cobbles up to 22.8 cm in diameter.
This very high–porosity material required numerous grout additives and
a speciic, tightly controlled work procedure to preclude excessive grout
loss. For each jet-grouted wall, interconnected soilcrete columns were
constructed with the double luid system to a depth of approximately 6
m along 243.8-m and 152.4-m stretches of esplanade, creating effective,
0.9-m-thick in-situ walls (Figure 6.51).
A very high–strength (20.68 MPa), corrosion-resistant soilcrete was
needed to meet speciication requirements. Extensive preconstruction
testing was therefore carried out to assess optimum mix design. Eleven
different mixes were tested, using a wide range of cement materials and
additives. During construction, numerous in-situ samples were retrieved at
close intervals at the interstice of soilcrete columns and tested for uncon-
ined compressive strength, continuity, and in-situ permeability. This
post-construction testing conirmed that the strength requirement in the
soilcrete walls had been achieved. Both phases of jet grouting were suc-
cessfully completed without detrimental impact to the park, the existing
structures, or the Hudson River.

REFERENCES

Abramovich, G.N. (1963). The Theory of Turbulent Jets, Cambridge, MA: MIT
Press.
Boehm, D.W. (2004). The utilization of jet grouting and soil mixing methods to
repair and support bulkhead structures, Proceedings of the American Society
of Civil Engineers Ports 2004 Conference, May 23–26, 2004, Houston, Texas,
United States.
Building and Civil Engineering Standards Committee, Technical Committee.
European Standard DIN EN 12716 (2001). CEN/TC 288, Execution of special
geotechnical works – Jet grouting, 05.18.07.
Burke, G.K, Cacoilo, D.M., and Chadwick, K.R. (2003). SuperJet grouting new
technology for in situ soil improvement, TRB Transportation Research Record
1721, Paper No. 00-0665.
 Jet grouting 257

Burke, G.K. (2004). Jet grouting system: advantages and disadvantages, Proceedings
of Sessions of the GeoSupport Conference: Innovation and Cooperation in the
Geo-Industry, Jan. 29–31, 2004, Orlando, Florida, United States; sponsored
by International Association of Foundation Drilling (ADSC) and the Geo-
Institute of the American Society of Civil Engineers.
Burke, G.K. (2007). Vertical and horizontal groundwater barriers using jet grout panels
and columns (GSP 168), Proceedings of Geo-Institute’s Geo-Denver 2007: New
Peaks in Geotechnics, February 18–21, 2007, Denver, Colorado, United States.
Burke, G.K. (2007). New methods for underpinning and earth retention, grouting
for ground improvement: Innovative concepts and applications (GSP 168),
Proceedings of Geo-Institute’s Geo-Denver 2007: New Peaks in Geotechnics,
February 18–21, 2007, Denver, Colorado, United States.
Burke, G.K. (2009). Quality control considerations for jet grouting, Geotechnical
News, December 2009, BiTech Publishers Ltd.
Burke, G.K. (2012). The state of the practice of jet grouting, Proceedings 4th
International Conference on Grouting and Deep Mixing, February 15–18,
2012, New Orleans, Louisiana, United States; sponsored by the Geo-Institute
of the American Society of Civil Engineers.
Carey, M.A., Fretwell, B.A., Mosley N.G., and Smith, J.W.N. (2002). Guidance on
the use of permeable reactive barriers for remediating contaminated groundwa-
ter, National Groundwater and Contaminated Land Centre Report NC/01/51,
Bristol, England: Environment Agency.
Camper, K.E. (2002). Safe passage: jet grouted columns on Rand & Bates tunnel,
Tunnel Business Magazine, December 2002.
Endo, W.K. (2009). The use of jet grouting techniques in permeable reactive barriers,
presented to Chemical Grouting Co., Ltd., Japan, July 29, 2009.
European Standard Amendment Draft 4093 to DIN EN 12716.
Franz, N.C. (1972). Fluid additives for improving high velocity jet cutting, Proceedings of
the 1st International Symposium on Jet Cutting Technology, British Hydromechanics
Research Association, April 5–7, 1972, Coventry, England, A7–93.
Fritsch, M. and Kirsch, F. (2002). Deterministic and probabilistic analysis of the soil
stability above jet grouting columns, 5th European Conference on Numerical
Methods in Geotechnical Engineering, NUMGE, Paris, Presses de LENPC.
Furth, A.J., Burke, G.K., and Deutsch, W.L. (1997). Use of jet grouting to cre-
ate a low permeability horizontal barrier below an incinerator ash landill,
Proceedings of the 1997 International Containment Technology Conference
and Exhibition, February 9–12, 1997, St. Petersburg, Florida, United States.
Hermans, J.J. (1953). Flow Properties of Disperse Systems, Amsterdam, The
Netherlands: North Holland Publishing Company.
Kirsch, F. and Sondermann, W. (2002). Zur GewÖlbestabilität über Soilcrete-
Korpern. 9. Darmstädter Geotechnik Kolloquium, Technische Universität
Darmstädt, March 14, 2002, Darmstadt, Germany.
Kirsch, K. (1997). Contractor’s view on the risks involved with deep excavations in water
bearing soils. Proceedings of the 14th International Conference on Soil Mechanics
and Foundation Engineering, September 6–12, 1997, Hamburg, Germany.
Lichtarowicz, A. (1995). Future of water jet technology basic research, Proceedings
of the 4th Paciic Rim International Conference on Water Jet Technology, April
20–22, 1995, Shimizu, Japan, pp. 13–26.
258 Ground improvement

Meinhard, K, Adam, D. and Lackner R. (2010). Temperature measurements to deter-

mine the diameter of jet-grouted columns, 11th DFI and EFFC International
Conference on Geotechnical Challenges in Urban Regeneration, May 26–28,
2010, London, England.
Noda, H., et al. (1996). Case of jet grouting for 10.8 m diameter shield. In: Yonekura
R., Shibazaki, M., and Terashi, M. (Eds.), Grouting and Deep Mixing,
Rotterdam, The Netherlands: Balkema.
Oakland, M.W., Ashe, J.M., Wheeler, J.R., and Blake, R.C. (2002). Design-build of a
jet grout access shaft for tunnel rehabilitation, Proceedings of North American
Tunnelling 2002, May 18–22, 2002, Seattle, Washington, United States.
Pollath, K. (2000). Baugrube Schleuse Uelzen II, Deutsche Gesellschaft für
Geotechnik. Baugrundtagung, September 18–21, 2000, Hanover, Germany.
Reichert, D., et al. (2002). Baugrube Domquarree. Deutsche Gesellschaft für
Geotechnik. Baugrundtagung, September 25–28, 2002, Mainz, Germany.
Shavlovsky, D.S. (1972). Hydrodynamics of high pressure ine continuous jets,
Proceedings of 1st International Symposium on Jet Cutting Technology, British
Hydromechanics Research Association, April 5–7, 1972, Coventry, England.
Shibazaki, M. (1996). State of the art grouting in Japan, Proceedings of the 2nd
International Conference on Ground Improvement Geosystems, May 14–17,
1996, Tokyo, Japan, pp. 857–862.
Shibazaki, M. and Ohta, S. (1982). A unique underpinning of soil solidiication utiliz-
ing super high pressure liquid jet, Proceedings of the Conference on Grouting
in Geotechnical Engineering, American Society of Civil Engineers, February
10–12, 1982, New Orleans, Louisiana, United States, pp. 685–689.
Shibazaki, M., Ohta, S. and Kubo, H. (1983). Jet Grouting, Tokyo, Japan: Kajima
Publishing Company, pp. 15–20.
Wachholz, T., et al. (2000). Planung, Konzeption der Schleuse Uelzen II und Berechnung
der Baugrube. Deutsche Gesellschaft für Geotechnik Baugrundtagung,
September 18–21, 2000, Hannover, Germany.
Wichter, L. and Kugler, M. (2002). Fehlstellen in DusenstrahlkÖrpern durch
Inhomogenita ten im Baugrund. Deutsche Gesellschaft für Geotechnik.
Baugrundtagung, September 25–28, 2002, Mainz, Germany.
Wittke, W. et al. (2000). Neues Konzept für die Schildeinfahrt in eine Baugrube im
Grundwasser. Taschenbuch für den Tunnelbau 2000, Essen, Germany: Verlag
Glückauf.
Yahiro, T. (1996). Water Jet Technology, Tokyo, Japan: Kajima Publishing Company,
pp. 15–29.
Yahiro, T., Yoshida, H. and Nishi, K. (1983). Soil Modiication Utilizing Water Jets,
Tokyo, Japan: Kajima Publishing Company, pp. 7–20.
Yanaida, K. and Ohashi, A. (1980). Flow characteristics of water jets in air, Proceedings
of the 5th International Symposium on Jet Cutting Technology, British Hydro-
mechanics Research Association, June 2–4, 1980, Hannover, Germany, A3–A33.
Yoshida, H. and Saito, K. (2009). Mechanism of cross jet and application, Proceedings
of the International Symposium on Ground Improvement 2009, December
9–11, 2009, Singapore, pp.159–164.
Yoshida, H. (2010). The progress of jet grouting in the last 10 years in the Japanese
market, Proceedings of the Deep Foundation Institute 35th Annual Conference,
October 12–15, 2010, Hollywood, California, United States, pp. 503–514.
Chapter 7

Soilfracture grouting
Eduard Falk and Clemens Kummerer

CONTENTS

7.1 Introduction ................................................................................. 260

7.1.1 Characteristics of fracture grouting ................................... 261
7.1.2 Construction and technical aspects ................................... 261
7.1.3 Basic applications of soilfracture grouting ......................... 263
7.1.4 Existing bibliography......................................................... 263
7.1.5 Basic project assumptions .................................................. 264
7.2 Technical developments ................................................................ 268
7.2.1 Further technical developments ......................................... 268
7.2.1.1 Settlement prediction and risk analysis................. 268
7.2.1.2 Sleeve pipe technology .......................................... 268
7.2.1.3 Soilfracturing combined with other methods ....... 269
7.2.1.4 Data processing .................................................... 269
7.3 Equipment .................................................................................... 270
7.3.1 Installation of grouting system........................................... 270
7.3.1.1 Drilling technology............................................... 270
7.3.1.2 Borehole setups..................................................... 271
7.3.2 Preparation of grout suspensions ....................................... 272
7.3.2.1 Storage system ...................................................... 272
7.3.2.2 Mixing technology ............................................... 273
7.3.2.3 Stored quantity of suspension ............................... 273
7.3.3 Grouting technology .......................................................... 273
7.3.3.1 Grouting pumps ................................................... 273
7.3.3.2 Sleeve pipes........................................................... 274
7.3.3.3 Packer system ....................................................... 274
7.3.3.4 Data recording...................................................... 276
7.4 Theory and design ........................................................................ 276
7.4.1 Soil improvement by hydraulic fracturing.......................... 278
7.4.2 Soil description .................................................................. 280
7.4.3 Mathematical models......................................................... 280
7.4.4 Monitoring ........................................................................ 282

259
260 Ground improvement

7.4.5 Basic information required for the design .......................... 282

7.4.6 Performance elements ........................................................ 283
7.5 Case histories and limits of application ........................................ 284
7.5.1 Settlement prevention ........................................................ 284
7.5.1.1 Increasing the load-bearing capacity of soils ........ 284
7.5.2 Passive settlement reduction............................................... 286
7.5.2.1 Deformation reduction for deep excavation pits ... 286
7.5.3 Back levelling of structures ................................................ 287
7.5.3.1 Stabilisation of a peat layer................................... 287
7.5.4 Concurrent and observational compensation grouting ...... 288
7.5.4.1 Compensation for settlements above
conventional underground excavations................. 288
7.5.4.2 Compensation for settlements above tunnels
excavated with tunnel-boring machines................ 289
7.5.5 Application limits .............................................................. 290
7.6 Test and control methods ............................................................. 292
7.6.1 Field trials.......................................................................... 292
7.6.2 Laboratory tests................................................................. 293
7.6.3 Monitoring technology ...................................................... 295
7.7 Environmental and contractual aspects ........................................ 295
References ............................................................................................. 296

7.1 INTRODUCTION

The phenomenon of hydraulic soil fracturing was initially observed as an

undesirable side effect of traditional grouting measures. The uncontrolled
propagation of soil fracs that were rapidly illed with grout suspension did
not achieve the objective of homogenously illed voids in granular soils. In
the past, high pumping rates that no longer allowed the grout to penetrate
the pore system of the soil continuously were regarded as a grouting defect.
It was crude oil technology using hydraulic soil fracturing for increasing
the permeability and thus the yield of oil ields that provided the impe-
tus for systematically applying geotechnical methods for using deliberately
produced voids in the soil. In the meantime, fracture grouting has been
used for systematically improving soil properties. The load-bearing capac-
ity and permeability of both granular and cohesive soils can be modiied
by incorporating a cement or solid matter skeleton. The repeated applica-
tion of this method also allows for the controlled raising of buildings with
very different foundation systems. The most spectacular use of the method
is found in connection with the complex tasks of compensating for settle-
ments which as a result of tunnel excavation threaten the structural integ-
rity of buildings above the tunnel.
 Soilfracture grouting 261

7.1.1 Characteristics of fracture grouting

High-viscosity grout is introduced through valves installed in the ground
in such a way that the sum of the reachable voids in the surrounding soil
accommodates only a small percentage of the amount of luid grout intro-
duced. As far as equipment is concerned, this method requires mixers for
producing suspensions rich in solid content and pumps that achieve a suf-
iciently high pressure increase of the suspension which is accumulating in
the soil. After the fracturing pressure in the soil has been exceeded, cracks
open up in the soil which are widened immediately by the subsequent grout.
By injecting small amounts of solid substance per grouting operation and
by repeatedly pressurising individual grouting valve positions, it is possible
to achieve a grout framework of hardened solid veins and lamellae (see
Figure 7.1).
The lamellae as described have an irregular shape and a median thick-
ness that can range from just a few millimetres to several centimetres. Soil
with large voids may need to be pre-treated in order to achieve a pressure
increase and, subsequently, to be able to carry out the process of hydraulic
fracturing. In order to control the development of ‘fracs’ in the ground
regarding their length, grout volumes are strictly limited and the low char-
acteristics of the grout are controlled by use of additives.

7.1.2 Construction and technical aspects

The use of hydraulic fracturing in construction technology has to meet
some important preconditions:

Figure 7.1 Supporting framework consisting of solid veins and lamellae for transferring
concentrated stresses.
262 Ground improvement

(1) Basic project assumptions: The achievable improvement of the soil

parameters has to be assessed just as realistically as the geometric
relations of foundations subjected to load, which allows for system-
atic lifting of a structure. Two- to ive-fold increases in stiffness can
be achieved.
(2) Performance description: The performance description on which the
works contract is based has to take into account realistic anticipated
progress, with observation periods and work interruptions necessar-
ily playing a part for organisational reasons.
(3) Time factor involved in the course of the construction work: A per-
manent soil improvement by hydraulic fracturing is achieved when
the number of grouting phases is large. Consequently, the desired
effect only occurs after the passage of a considerable amount of
working time. For compensation grouting it is particularly important
to include in the time schedule a suitable period for installing the
grouting system and pre-treating the soil up to the point in time when
the structure is ready to be lifted.
(4) Measuring technology: The method is controlled by means of
well-developed measuring technology, which makes it possible
to observe both surface movements as well as any deformation in
the subsoil. The reliability of the measured values and their early
evaluation are essential preconditions for a successful application
of the method.
(5) With hydraulic fracturing the soil is improved in small stages and
with the objective of achieving a permanent increase in the lateral
soil resistance. Consequently, only those tests which, both in terms
of time and space, relect the geometry of the actual application can
provide conclusive evidence of performance.
(6) Application limits: After very promising results were achieved in con-
nection with raising buildings and compensating for settlements
during the last three decades, there exists a tendency to exceed pre-
viously known application limits. However, an essential element of a
successful application is observing maximum injection rates per soil
unit and working day. As far as economic aspects are concerned,
it has to be taken into account that a sensible decision on the use-
fulness of the application can be taken only if the prevented dam-
age and its frequently complex consequential effects are realistically
assessed.

The purpose of this chapter is to make available relevant principles on

which decision making can be based and which can be adapted to the
respective speciic project conditions. The following examples should be
regarded as an overview of those areas of application where hydraulic frac-
turing has been used in recent years.
 Soilfracture grouting 263

7.1.3 Basic applications of soilfracture grouting

(1) Modiication of soil stiffness, such as soil homogenisation underneath
a machine foundation (see Figure 7.2).
(2) Stabilisation of long-term or current differential settlements. For exam-
ple, differential settlement of a church which had continued over centu-
ries and eventually became a safety problem (see Figures 7.3 and 7.4).
(3) Compensation of existing settlements, such as levelling of structures
affected by settlements in the vicinity of deep excavation pits or tun-
nel excavations (see Figure 7.5).
(4) Active compensation of absolute and/or differential settlements as
they occur. For example, compensating for settlements above cavity
structures, which are often built in several phases by means of tradi-
tional excavation (see Figures 7.6, 7.7, and 7.8).
(5) Conditioning and compensation for settlements that occurred as a result
of excavation with tunnel-boring machines (see Figures 7.9 and 7.10).

7.1.4 Existing bibliography

It should be noted that because of the spectacular results, many cases of the
application of compensation grouting have found access to international

Heavy tool processing machine

~18 m

+/– 0.0
Reinforced slab

Non homogeneous Criterion:

fill ‘the dynamical elasticity
modulus of the ground’
6.0

Sandy gravel Zone of treatment

Figure 7.2 Consolidation of a demolished, inadequately reilled basement underneath

the foundation of a precision machine, with the objective of homogenising the
dynamic deformation modulus of the soil.
264 Ground improvement

Clayey and sandy silt,

with water-bearing sand layers in humid seasons

Figure 7.3 Shrinkage processes caused by cyclic drying of the soil lead to differential
settlement occurring in stages.

Longitudinal section Plan view

Intensity of
treatment 1

45 m
Intensity of
treatment 2
10 m

26.4 m
Intensity of treatment 1 Intensity of treatment 2

Figure 7.4 The intensity of the stabilisation process by hydraulic fracturing to be adapted
to the existing loads.

technical conferences, but the basic project assumptions still are a subject
of discussion by experts. This is why the attached references contain publi-
cations that deal with the further development of mathematical models, in
particular involving numerical methods, current improvements in measur-
ing technology, and well-founded views on the interaction resulting from
forced deformation between the soil and the structure.

7.1.5 Basic project assumptions

In principle, it has to be stressed that hydraulic fracturing consists of
imposing soil deformation by injecting grout rich in solid components.
 Soilfracture grouting 265

~30 m

+/– 0.0

–6.0
g’
epin
‘Cre
8m

Clayey silt
Soilfrac treatment

Weathered rock

Figure 7.5 Stopping creep movements after deformation in the retaining structures had
led to cracks in existing structures.

[m]
55
Historical buildings
50
45

40 Working
shaft Historical buildings
35
30

25 Soilfrac

20

15

10

5
0

–5
–10

0 10 20 30 40 50 60 70 80 [m]

Figure 7.6 Compensating for settlements above two station buildings for an underground
railway which was built in stages over a period of approximately two years.
266 Ground improvement

Rua Garrett
Rua capelo

Shaft 1 Station tunnel west

Shaft 2
Station tunnel east Shaft 3

Rua lvens

Rua nova do almada

Figure 7.7 Arrangement of shafts and drilling for the installation of grouting pipes for
settlement correction in densely built-up area of approximately 15,000 m2.

[mm]

20

–20
Settlement

Settlement without compensation

–40
Settlement with compensation
–60

–80

–100

–120

17 Dec 5 Feb 26 Mar 15 May 4 Jul 23 Aug 12 Oct 1 Dec

Date

Phase 1
Phase 2, 3
Phase 4, 5
Phase 6, 7, 8, 9

Figure 7.8 Illustration of the average compensation success, indicating the compensation
procedure in small stages during the individual excavation phases.
 Soilfracture grouting 267

Surface

Fill 10 m
11.5
Sand
18 m
1.5
Sleeve pipes
5.0
Gravel

Tunnel 9.4 m
in diameter

Figure 7.9 Compensation grouting for the protection of a railway bridge with 9.4-m twin
tunnel excavation and differential settlement limits of 1/3,000.

900 30
TBM 1 TBM 2
158 m3 67 m3
Total grouted volume [m3]

Daily grouted volume [m3]

800 20

700 10

600 0
20/06 04/07 18/07 01/08 15/08
2005 2005 2005 2005 2005

Figure 7.10 Grouted quantities for irst and second TBM drive. The mean advancement
rate was 18 m/day.

The criterion for effectively applying the method consists of being able
to control the movements in the soil as well as the interaction between
such movements and the structure concerned. The more or less com-
plex requirements demand close interdisciplinary cooperation between
geotechnology, structural analysis, structural process technology, and
268 Ground improvement

measuring technology. A pre-condition for effectively applying the

method is the evaluation of data of different disciplines being compiled
by a structural engineer competent in these disciplines and capable of
making decisions within the framework of the application of an obser-
vational method.

7.2 TECHNICAL DEVELOPMENTS

After it was recognised that the phenomenon from traditional grouting and
the equipment in the oil ields can be combined successfully to solve geo-
technical problems, the lifting method using cement grout was irst used in
Essen for the purpose of raising a coke furnace, as described by Bernatzik
(1951). Essential steps for improving the method and for widening its range
of application consisted of adapting the valve pipe technology (in approxi-
mately 1970) and in integrating electronic data processing, which allowed
for actively compensating settlement troughs.
In the meantime, the method has spread to many geographic locations.
It is known to have been used in Germany, Italy, Austria, The Netherlands,
Portugal, Spain, Belgium, United Kingdom, United States, Canada,
Australia, and Puerto Rico.

7.2.1 Further technical developments

7.2.1.1 Settlement prediction and risk analysis
Ever since experience with inner-city tunnel construction in recent
decades highlighted the considerable extent to which environmental dam-
age can inluence the total costs of such measures, systematic settlement
analysis has occupied an important place within overall project planning.
This includes a detailed examination of the condition of existing struc-
tures in the area at risk and an evaluation of possible damage within the
framework of a comprehensive risk analysis. On the basis of these data
the decision has to be made which stages or areas additional measures
can be economically added to the settlement-reducing measures in tunnel
excavation itself.

7.2.1.2 Sleeve pipe technology

The improvement in the grouting system consisting of long-life valve pipes
and double packers made it possible to grout individual sleeves many times;
to use long pipe lengths and, if necessary, to carry out controlled drilling
operations. High-quality pump control devices do not only allow a large
number of decisions to be made on parameters, but also enable these data
 Soilfracture grouting 269

3m

3m

O
3m

L
Sleeves at
0.5 m spacing

Figure 7.11 Sleeve pipes can be arranged on several levels and can serve different
purposes. This igure shows a detail of a lifting mat used in compensation
grouting with two horizons: ‘O’ and ‘L’ for the purpose of stress distribution,
plus a central lifting layer ‘M’.

to be recorded and automatically presented. High-performance drilling

methods and the use of lexible drilling assemblies gives the option of the
use of shafts and also the adaptation of existing working areas such as
excavations or tunnels, which permit the method to be used in conined
inner-city spaces as well (Figure 7.11).

7.2.1.3 Soilfracturing combined with other methods

To meet dificult requirements and to ind economical solutions, the method
has already been developed and used in combination with several geotech-
nical methods: compaction grouting, jet grouting, pipe roofs, hydraulic
jacking, loating pile foundations.

7.2.1.4 Data processing

Recording and storing of data obtained in grouting itself and from the
monitoring-based observations are not suficient for ensuring a profes-
sional application of the method. Only the use of professional visualisa-
tion programmes and the combination of data by means of individual
software modules makes it possible for the site manager to take real-time
decisions on grouting programmes that have to be modiied continuously
(Figure 7.12).
270 Ground improvement

Silos
Cement Filler Bentonite

Pump container
High shear Agitator
mixing tanks Pumps Control unit
Water

Measured data

Grouting programme

Data aquistion +
Monitoring
site management

Zone of treatment

Measured data

Figure 7.12 Data on deformation measurements and parameter recording are used by
the site manager for determining actual grouting programmes.

7.3 EQUIPMENT

7.3.1 Installation of grouting system

7.3.1.1 Drilling technology
The sleeve pipes (also: Tube à Manchettes, TAM) can be installed in bore-
holes stabilized with mud-lush or by a casing. Both the rotary drilling
method using a down-the-hole hammer and methods driving the casing
with an external hammer are used. In soft stable soils it is also possible
to apply auger drilling methods. The resulting spoil material is conveyed
by air pressure or water lushing. In special cases it is necessary to use
directional drilling where long bores with limited deviation tolerances are
required. Moreover, directional drilling enables curved drilling. The curva-
ture that can be achieved must be larger than 120 m.
 Soilfracture grouting 271

Grouting

Surface

Detail Detail

TAM Annular grout Double packer

Figure 7.13 Sleeve pipes (TAMs) are installed in a stable (cased) borehole and sealed by
an annular grout. Individual valves can be pressurised by double packers.

In the case of double-packer technology, it is advisable to limit the pipe

length to approximately 50 m, although borehole lengths exceeding 75
m have been successfully achieved. It should be noted that in the case of
extended drilling length the general application problems increase dispro-
portionally. Directional drilling can be used for these cases to limit drilling
deviation. In the case of controlled 100 m boreholes, the double-packer
technology can be replaced by ‘no-return’ valve technology (Figure 7.13).

7.3.1.2 Borehole setups

Eficient drilling masts for producing boreholes in excess of 20 m comprise
lengths of at least 3 m, but preferably approximately 5 m. Any drilling
shaft, trench/excavation, and working tunnel should be designed to be at
least 1 m wider than the length of the drilling masts. Arrangements for
orderly discharge of lushing water containing cement have to be included
in the works schedule, just like the safe introduction and lifting of the drill-
ing and grouting equipment. All safety requirements have to be observed
during the different operating phases and are normally combined in a safety
schedule which is pointed out to those participating in the project prior to
the commencement of work. If shafts have to be located in the direct vicin-
ity of the area where the lifting operations take place, it may be necessary
to provide a dilatation joint, if necessary even underneath the water table
(Figure 7.14). Typical drilling equipment for a shaft is shown in Figure 7.15.
272 Ground improvement

3. From surface

1. Shaft 2. Basement
4. Excavation pit

Soilfrac treatment
Tunnel
5. Service tunnel

Figure 7.14 As far as the drilling geometry is concerned, a large number of options are
available for reaching distinct areas in the soil to be treated.

Figure 7.15 Drilling rig for a 3.5-m diameter shaft.

7.3.2 Preparation of grout suspensions

7.3.2.1 Storage system
The storage system must facilitate separate storage of the major compo-
nents required for mixing suspensions. It is necessary to provide steel
or plastic tanks for storing water in drinking-water quality, silos with a
capacity of at least 20 tonnes for bonding agents (cement, lime, ly ash)
and iller material (limestone, slag, bentonite) as well as small containers
for additives.
 Soilfracture grouting 273

[%]

100%
= 1m3 w w w w w

80
Components:

c w - Water
c
60 c
% of weight

c c - Cement

c f - Filler

40 s - Sand
f f b - Bentonite
f
20
f
s

b b b b b

M1 M2 M3 M4 M5

Figure 7.16 Examples of suspension compositions that can be used in different soils and
operations.

7.3.2.2 Mixing technology

It involves the use of so-called colloidal mixers or high-frequency mixers
which allow the homogeneous mixing of suspensions with a water/solid
ratio of 0.45. Even with grouts with a high solid content and bentonite
added, the mixing capacity has to ensure an adequate supply of material to
the proposed number of pumps (Figure 7.16).

7.3.2.3 Stored quantity of suspension

Limited quantities of 500 – 1000 litres are stirred in agitator tanks with
electronic control devices (‘multirangers’) to ensure that grouting units are
supplied continuously and that larger suspension quantities are prevented
from being ‘stirred dead’ due to any interruption in the grouting operation.

7.3.3 Grouting technology

7.3.3.1 Grouting pumps
Modern grouting pumps provide different control options, thus permit-
ting an automatic reaction on the development of pressure at a constant
274 Ground improvement

[L/min] [bar]
Grout pressure
60
15

40
Pumping rate

10
Pressure
Pumping rate
(const.)
20
5
Pumped volume Σ 20-80 I

Duration of one grouting phase [min]

Figure 7.17 Typical injection graph showing a drop in pressure after the occurrence of
‘fracs’.

or variable pump rate. The pumps have to be suitable for a pressure range
of 0–100 bar and a pump rate of 1–20 litres per minute; any parameter set
within these ranges is expected to be kept constant even for suspensions
with a high solid content. In modern grout modules, it is a common practice
to combine 2–8 pumps of similar design. The essential pump parameters
are either graphically recorded immediately or stored electronically and, via
software programmes processed in databanks and printed out. In any case,
the type of data saved has to ensure that in each individual grouting opera-
tion the pressure and quantity ratios must be clearly associated with the
respective location of the grouting operation. This is the reason why modern
grouting data recording and visualisation programmes are coupled; they
permit an early interpretation of data (Figures 7.17, 7.18, and 7.19).

7.3.3.2 Sleeve pipes

Pipes typically available consist of PVC, ibre glass, or steel. Their diameter
ranges from 1–4 inches and the distance between valves amounts is 0.33–1 m.
In special cases, the steel pipes are reinforced and the rubber valves are
protected by steel rings or special covers in such a way that either the pipes
can be driven directly or that the function of the reinforcement of the soil,
for example the shape of the pipe umbrella, is supported. In principle, by
surrounding the grouting pipes with a so-called skin-forming grout (stable,
low-strength, but stiffened sleeve grout) a direct connection between the
individual grouting valves is prevented.

7.3.3.3 Packer system

Double packers are used to limit the grouting operation to one speciic
valve of the grouting pipe (Figure 7.20). The packer elements consist of
 Soilfracture grouting 275

Vr

lFrac
Q = x l/min

b Fr
V=t.Q

ac
daverage

Mm

AFrac = lFrac . bFrac lFrac = f {Q, V}

VFrac = AFrac . daverage daverage = 2 ÷ 30 mm

Figure 7.18 A frac usually has a lat, oblong shape and starts from an initial crack in the
sleeve grout. However, frequently there occur secondary fracs whose geo-
metric description can be given in statistical form only.

a wire mesh and have a rubber surface, and are inlated either with com-
pressed air or water.
The pressure applied has to be clearly higher than the maximum
grouting pressure expected. After completion of the grouting opera-
tion, the grouting pipes are cleaned by high-pressure hoses integrated
into the packer system or by separate cleaning systems. The quality of
the individual components and their careful use are the key to an effi-
cient process application and their successful use over longer periods
of time.

Figure 7.19 Coloured frac sample deriving from grouting in cohesive soils.
276 Ground improvement

Figure 7.20 Double packer for 2” grouting pipes.

7.3.3.4 Data recording

A compensation grouting operation generates a huge amount of data, which
has to be suitably administered and made available to the site manage-
ment in order to make appropriate decisions. For instance, there have been
applications in the past with more than 100.000 individual grouting opera-
tions and the continuously measured values of more than 100 measuring
elements had to be recorded over a period of more than one year.
Associating the data in terms of time and place is just as important as
combining the effects, which should start from the excavation stages—
from the grouting measures and the inluence of individual regions of the
structures positioned above. The objective is therefore to generate a simpli-
ied presentation of the structures and the soil in a form that shows the
interaction between excavation, grouting operation, and structure. When
designing the measuring system, care has to be taken to ensure that the
actual measuring accuracy and the frequency of data recording indicate
movement tendencies. Important external inluences such as temperature
variation have to be iltered out.
Valuable tools in the application of compensation grouting are graphical
display systems which, for example, at any time show the development of
pressure ratios during grouting or the distribution of quantities inside geo-
metrically limited units.

7.4 THEORY AND DESIGN

Injecting solid material into the soil leads to deformation on all sides. The
directions in which individual injections spread largely depend on the
homogeneity conditions in the soil. From a statistical point of view, it has
to be assumed that the greatest part of the volume introduced into the soil
 Soilfracture grouting 277

leads to deformations whose amounts are distributed proportionally rela-

tive to the respective stress conditions. Only a small part of the movement
rates is caused by the compression of the existing soil, as the highest effec-
tive injection pressure cannot greatly exceed the magnitude of the highest
existing standard stress. However, locally and over a short period of time,
higher forces can become effective if tensile forces within the grout skeleton
which have already been set and cohesion forces in the soil are activated
(Figure 7.21).
Experience has shown that if a limit rate with respect to the injected
quantity per unit time within a limited soil volume is exceeded, said forces
are overcome and, in consequence, the time-dependent deformation resis-
tance is clearly reduced. This conclusion is based on the observation that
the eficiency of injected quantities is reduced in the course of a lifting
operation, if the quantity injected per working day is increased excessively
in order to try to achieve a greater lifting speed. The above-mentioned limit
balance between effective injection pressures applied (not the injection
pressure measured at the pump) and the annularly acting pressure forces
in the surrounding soil depends on the respective soil characteristics to
such an extent that global recommended values for reliable injection quan-
tities cannot reasonably be given. It is advisable to monitor limited injection
areas by making use of the available measuring technology and to deter-
mine the achievable ‘grout eficiency factor’ (deined as the ratio between
‘heave volume’ monitored on surface to grouted volume) and lifting speed
by varying the injection parameters (Figure 7.22).
Clear information on the deformation rates and their directions as a
result of the injection of solids cannot reliably be obtained by only observ-
ing surfaces. Information from the soil is essential, with extensometers and
inclinometers able to provide useful service. For the purpose of checking

Fracture grouting

Figure 7.21 The model of the ‘conining’ ring comprises the sum of all forces that allow
for a central lifting injection.
278 Ground improvement

Phase Horizontal (H) Vertical (V) V

Pre-treatment 100% 0%

Multi-stage injection 95 ÷ 100% 0 ÷ 5% H

before actual heave
H
Heaving phase 75 ÷ 95% 5 ÷ 25%

VHeave = (0.05 ÷ 0.25) x Vinj.

V Heave
σoverburden + external load σoverburden + external load

Displacement

σ1 σ1 σ1 σ1
σ2,3 ≤ σ1 σ2,3 ≤ σ1 σ2,3 ≥ σ1 σ2,3 ≥ σ1

Resistance

σoverburden + external load σoverburden + external load

Figure 7.22 Heterogeneities in the soil and the stress distribution inluence the defor-
mation direction, which is the result of an imposed addition of material and
all-round resistance.

the accurate arrangement of geotechnical measuring instruments, it is con-

servative to follow the rule that almost all quantities of solid material intro-
duced into the soil have to be identiiable as deformations in the soil.

7.4.1 Soil improvement by hydraulic fracturing

A permanent improvement in cohesive soils is achieved by producing a
continuous supporting framework consisting of a hardened solid skel-
eton. Since it is only the homogenisation of stress conditions in the soil
and the closure of the solid skeleton that permits a supporting effect
independent of the original soil, the improvement curve of a multiphase
injection application is not to be regarded as being linear by any means.
 Soilfracture grouting 279

The stresses acting on foundations of up to approximately 1000 kN/m2

typically occurring in structures can often only be transferred into the
ground if soils are improved (e.g., by soil fracturing). It is possible to
transfer stresses in the range of 2.0–3.0 MN/m2 with negligible deforma-
tion rates implementing fracture grouting measures. Borderline cases are
soft soils. Whereas effective consolidation has already been achieved in
peat layers, there are no examples yet for very soft and structureless types
of soil, although in such cases, the selection of a suitable injection agent
promises success, too.
Soil improvements by hydraulic fracturing are always carried out with
the intention of generating a controlled stress low in the ground. Because
of the low material-strength values of the soil and the intention to include
said soil in a supporting system, it is frequently necessary to increase the
initial soil stiffness by 2–5 times the existing value.
As the deformation method described involves the displacement of large
soil masses by small individual amounts, the respective plans must take
account of the principle that slim deformation elements should not be con-
sidered in planning such methods. Figure 7.23 shows basic geometric rela-
tions that result from experience with different lifting operations. In such
considerations the different load intensities of adjacent foundation inter-
faces and the different depths at which they are located also play important
parts.

A
Reference level
σA B
σB
bA

αlimit bB
αA αB tinj

dA dB

hinj
rinj

Figure 7.23 Recommendations for geometrical relations and load intensities to permit
controlled levelling: αlimit = 10–25°; dA /dB ≤ 1.5 with αA + αB ≤ 20°; bA /dA ≥ 1
resp. bB/dB ≤ 2; σA /σB ≤ 3 with αA + αB ≥ 20°. Situations exceeding these
indications need additional veriication or practical trial.
280 Ground improvement

Table 7.1 List of parameters with signiicance for project considerations

Designation Short symbol Unit Importance
Grain-size distribution – mm 1
Coeficient of non-uniformity U – 1
Moisture content W – 1
Porosity n – 1
Void ratio e – 1
Relative density D – 1
Consistency index Ic – 1
Liquidity index IL – 1
Coeficient of permeability k m/s 2
Young’s modulus E kN/m2 2
Shear strength parameters φ’, c’, φr °, kN/m2,° 2
Undrained shear strength Cu kN/m2 3
Model of rheology 3
1 – Absolutely necessary
2 – Important for calculations
3 – Signiicant for numerical models

7.4.2 Soil description

In principle, hydraulic fracturing is suitable for improving all types of soil
with an adequate consistency. However, to be able to quantify the require-
ments which have to be met, it is necessary to have speciic information on
the initial soil properties. Table 7.1 contains a selection of the necessary
parameters and an evaluation of the effect on the quality of the prediction.
In principle, it has to be said that such a method is used almost exclusively in
dificult nonhomogeneous and anisotropic soil conditions. Even if it would
be possible to describe accurately individual soil layers, it would be almost
impossible to describe the problems generated by the layering effect and
by even greater irregularities such as karst illings. However, it is precisely
these irregularities that usually are the actual causes for the occurrence of
considerable differential settlements. Therefore, the most important objec-
tive of the project preparation phase is recognising the nature of the causes
of differential settlements and formulating a concept regarding necessary
further information during the execution of the work.

7.4.3 Mathematical models

Simple evaluations of the geometric relationships occurring in the course of
lifting injections can be made by using a ‘block model’ in which the zones
of different treatment intensities are modelled in the form of an idealised
‘conining’ stress ring, and the lifted zone in the form of a centrally posi-
tioned lifted piston.
 Soilfracture grouting 281

The presently available numerical calculation methods allow the model-

ling of many deformation phenomena in the soil and their interactions with
structures positioned above. In the past, a considerable impediment regard-
ing the simulation of the lifting injections has been found by the continu-
ous change of the input parameters as the soil properties change during the
treatment of the soil.
Recent publications describe two- and three-dimensional analytical and
numerical solutions for modelling the effect of compensation grouting.
With these models, case histories have been investigated matching the real
situation in a considerable manner.
A back analysis of fracture grouting applications was presented by
Schweiger et al. (2004). The numerical model adopted simpliied analytical
and numerical models according to the inite element method for the settle-
ment correction simulation for the central station of Antwerp. The Finite
Element Mesh for the calculation is represented in Figure 7.24.
With the proposed model, it was shown in a reasonable manner that
FEM calculation can be valid for the prediction of the compensation effect
(Figure 7.25). However, the prediction of exact grouting quantites is still
dificult as numerical models do not take into account various effects of
the practical grouting behaviour (illing of voids, variable grout eficiency,
effects of the grouting parameters, and so on).

Levelling point Levelling point

P455 P470

Foundation

Grouting zone

Pipes A-B-C-D

Tunnel

Figure 7.24 Plane strain Finite Element model for the back analysis of the construction
stages for the tunnelling at Antwerp Central station.
282 Ground improvement

4
Vertical displacement [mm]

1 Measurement P470
Calculation P470

0
g

B*

*
D

C*

C
t

v
in

ca
pe
ro

pe
pe

p
nn

pe
p

ex
p

m
m
tg

Pi

Pi
m

Pi

m
gi

Pi

Co
Co

el
Co
ac

Co
Be

nn
nt

Tu
Co

Figure 7.25 Comparison between ield measurements and FEM back analysis for the
compensation grouting work of the Antwerp Central station.

7.4.4 Monitoring
The concept for comprehensive deformation measurements always
includes the two completely different sets of problems of the structure
on the one hand and of the soil underneath on the other hand. As far as
structures are concerned, monitoring the level of supporting components
is of primary importance. In addition, loors and existing cracks can
be provided with instruments whose values are read automatically or
visually. Measuring instruments such as inclinometers, extensometers,
incremental extensometers, settlement piezometers, and similar devices
can be arranged in the ground with the objective of determining spatial
deformations and their directions in the direct vicinity of the area treated
(Figures 7.26 and 7.27).

7.4.5 Basic information required for the design

To be able to design a project involving the method of hydraulic fracturing,
it is necessary to make reliable assumptions regarding

r Expected settlements, when excavations are encountered

r Tolerable settlements
r Grouting eficiency

Moreover, the pretreatment phase, where the voids are illed in order to
condition the soil, and preheaving have to be taken into account.
 Soilfracture grouting 283

20 m
15
10
5
-1.1
1.0

1.4
1.7

0.1
1.4
0.0

0
1.5

0.00
1.5
1.00

0
1.00

-2.00
00

-3.2
-1.0

230
0.3
0.8
1.1

1.

-0.6
2.00
1.00

0
0

2.0
1.0
0.4

0.00

0
1.1

-1.0
0

0
3.0

-2.0
0
2.0 4.00
1.2

0.0
1.00
0.00

0
-0.7

2.8

4.8

2.0

0
0

1.2
5.0

-3.0
4.4
3

5.1
1.6

0
2.7

1.7
4.0

0
4.4

1.0
4.00

-1.00
2.00

-2.00
3.00
5.00

-3.00 -3.4
-2.4
-0.5
1.00

0
2.0

0
5.0
1.9
0.1

0
0.9

3.0

4.00

0.00
4.0

0
1.0

-0.4
3.00
2.0
1.5

0
1.6 2.0

-1.00

-2.00
4.3 4.4

3.4 3.7
1.7

4.3

0.1
0.0
0

3.003.7
-0.1

1.0

2.00
1.0

0.0-0.7
0
2.00

0
3.0
1.1
2.5

-3.0
-0.4

2.5

2.4
2.1
3.4
2.0

-0.9
1.5

1.1
3.6 3.1
2.9

1.00

0
-2.0
2.00

3.00
0

2.3

0.5
2.6

2.2
1.0

1.7

-
9.9

1.9
-.9
465

Figure 7.26 Example illustrating the deformation of a historic railway station which was
continuously monitored by means of a complex measuring system.

7.4.6 Performance elements

The plan for the application of hydraulic fracturing is normally followed by
a performance description which is used as a basis for the works contract
to be concluded with the contractor. Table 7.2 contains a list of normally
occurring working phases and a selection of necessary speciications. In
addition, there are proposed units according to which the very different
services have to be evaluated, depending on the type of project.
284 Ground improvement

Grouting area

Surface

s
∆s
Volume to be compensated
Settlement curve

s...Allowable absolute settlement

∆s...Allowable differential settlement

Figure 7.27 Actual settlements and the allowable deformation state basically determine
the volume to be considered during the compensation phase.

When applying soil fracturing in connection with very complex infra-

structure measures, it may be advisable to include in the performance
description regularly occurring interruptions in the construction work; in
addition, a distinction has to be made between phases when the equipment
is ready for injection and phases of pure measuring technical observation
and equipment idle time, respectively.

7.5 CASE HISTORIES AND LIMITS OF APPLICATION

Hydraulic fracturing can be used for different objectives and types of settle-
ment reduction operations. The spectrum ranges from pure prevention or
repair works after settlement damages have already taken place to simulta-
neous compensation of excavation induced settlements for inner-city areas.
A number of examples illustrate the opportunities for this method and
should provide creative ideas for the planning engineer.

7.5.1 Settlement prevention

Soil improvement by hydraulic fracturing takes place to homogenise the
stress conditions in large soil volumes or to increase the load-bearing capac-
ity of soils to enable them to support additional loads. It is important to
point out the economic nature of this method compared to deep foundations
systems because a suficient degree of soil improvement has been reached for
the load acting on the soil-structure interface when initial heave is achieved.

7.5.1.1 Increasing the load-bearing capacity of soils

In the case of already over-loaded strip foundations of a three-story build-
ing it was decided to improve the soil down to 8 m depth to enable the
 Soilfracture grouting 285

Table 7.2 Performance elements of compensation grouting

Performance element Important speciications
Licences for drilling areas and Duration of use/number of measuring points to be
construction site equipment, provided
access to structures for
exploration purposes
Exploring installation situation Type of exploration
Safety measures
Recording details of structures Type of structure
Preparing and maintaining drilling Shaft depth and diameter
surfaces Types of excavation/expenditure for adapting
existing spaces
Access situation
Setting up drilling equipment Mean and maximum drilling length
Drilling method, borehole diameter
Diameter of valve pipe to be built in, spaces
available
Execution of drilling operations Borehole diameter, diameter and wall thickness of
installation of sleeve pipes pipes, material properties, distance between valves,
valve characteristics
Filling the annular space Suspension mixture data
First injection Number and sequence of individual injections,
quantity injected per injection operation
Mobilisation of grout station Number of injection units to be operated
separately
Multiple injection; Number of injection operations and quantity to be
preconsolidation and introduced per injection operation, minimum and
‘conditioning’ maximum pump rates
Prelifting and injection up to Number of injection operations and quantity to be
point when the structure is introduced per injection operation, minimum and
ready to lift the structure, maximum pump rates
‘contact’
Lifting ‘compensation injection’ Listing lifting stages and lowest and highest
settlement rates to be compensated for per shift
Installing an automatic measuring Type and number of measuring elements to be
system installed
Providing and maintaining the Assessing the required functional periods, required
measuring system for the accuracies, measuring frequencies
duration of the project and/or
for the period of continued
observation
Data recording and visualisation Providing, in a standard form, the injection
parameters and data from monitoring, frequency
and type of data sets to be handed over
Technical site management Interpreting the measured data jointly with
designer, setting up individual injection
programmes/evidence of experience with similar
construction situations
286 Ground improvement

Additional load due to added flats

Small
Surface strip foundations

Soft silt

Stiff silt
Zone of improvement

Figure 7.28 Increasing the load-bearing capacity in the areas where additional loads have
to be carried due to addition of two storeys.

building to support, free of undue settlements, additional loads resulting

from the construction of additional storeys (Figure 7.28).

7.5.2 Passive settlement reduction

If differential deformations are expected, a pretreatment with the objective
of improving the soil can considerably reduce the absolute extent of such a
deformation and, if necessary, allows an active correction of actually occur-
ring movements. Normally, the costs of such a measure are signiicantly lower
if the installation of the sleeve pipes is already included in the construction
design.

7.5.2.1 Deformation reduction for deep excavation pits

The most economical variant of securing a construction pit by means of
an anchored diaphragm wall was likely to result in an extent of horizon-
tal deformation that was unacceptable for the adjacent buildings. As an
alternative to secondary underpinning of foundations by jet grouting, it
 Soilfracture grouting 287

Historical building

10 m
Wall displacement

Settlement prediction
15 m

Excavation
level

Zone of improvement
Sand
Sealing slab

Figure 7.29 Passive settlement reduction by carrying out speciic soil improvement in
the area subject to deformation.

was considered to improve the block of soil behind the diaphragm wall
by means of soil fracturing which, if needed, would also allow for volume
losses to be compensated during the excavation phase (Figure 7.29).

7.5.3 Back levelling of structures

There are several traditional methods for performing repairs of settlement-
damaged buildings, including the use of piling systems, jet grouting, and hand
digging. Most methods have complications connected with the use of working
areas inside the building. In addition, lifting of buildings by hydraulic jacking
is a very complicated operation that even if applied accurately requires a very
sophisticated control technology and a suficiently intact and stable structure.
By adapting the soil fracturing technology to small-scale applications, it is
possible to obtain a technically and economically interesting alternative and
the access to the building is limited for observation purposes only.

7.5.3.1 Stabilisation of a peat layer

A wedge-shaped layer of peat existing underneath the foundation slab
has not been properly identiied during the construction of a multi-storey
288 Ground improvement

Surface

10 m

1:85 Fill/gravel

Peat
Lfting zone
Moraine
Zone of stabilization

Figure 7.30 Lifting an apartment building.

building and as a consequence a shallow foundation with a concrete slab was

executed. The resulting differential settlements led to a maximum inclination
of 1:85, and thus the serviceability of the building was questionable. On this
occasion it was successfully shown that for the irst time organic soil can also
be improved by hydraulic fracturing, although at higher costs (Figure 7.30).

7.5.4 Concurrent and observational

compensation grouting
Ever since electronic measuring systems have allowed a real-time deforma-
tion measurement, it is possible to actively compensate for concurrently
developing settlement troughs by means of soil fracturing. In this context,
it is essential to assess realistically the total settlement to be expected as
well as the highest settlements occurring during a limited period of time.
Obviously, minimising settlements by excavating a tunnel in a suitable
manner constitutes the essential element of an environmentally acceptable
construction method in inner-city areas.

7.5.4.1 Compensation for settlements above

conventional underground excavations
Over a period of 18 months, historic structures had to be protected from
differential settlements caused by station tunnels with internal diameters
of up to 20 m. Total settlements in excess of 100 mm were compensated
in lifting stages of 2 mm per lifting operation, so that the position of the
 Soilfracture grouting 289

Park garage

0 Shaft
Silt

Grouting pipes
Marl
Tunnel

10 m

Figure 7.31 Settlement compensation for a car park with very small cover.

building foundations prior to the tunnel driving operation remained almost

unchanged (Figures 7.6, 7.7, and 7.8).
Another case study was the excavation of a large tunnel underneath a car
park where the cover between the foundation and the shotcrete lining had
a minimum distance of less than 1 m. The 45-m-long grouting pipes were
installed from a shaft inside the building without interference by means of
the grouting operation the tunnel excavation was performed with limited
settlements of about 7 mm (Figure 7.31).

7.5.4.2 Compensation for settlements above tunnels

excavated with tunnel-boring machines
Excavating tunnels with tunnel-boring machines differs fundamentally
from using mining methods, mainly with respect to time characteristics.
Because of higher driving speeds up to 25 m/day and occasionally abrupt
interferences in the zone of the earth pressure support, greater attention
has to be paid to passive effect of compensation grouting. As a result of
the number and types of sleeve pipes, it is possible to additionally use their
reinforcing effect or the soil. Furthermore, an extensive monitoring system
extended over a suficient area contributes considerably towards control-
ling shield machines when installing a suitable communication system.
One of the major compensation grouting jobs in Europe was the protec-
tion of approx. 22,000 m2 of urban area in the centre of Leipzig, Germany.
The twin tunnel with a diameter of 9 m had a cover ranging from 7.5 to
15.6 m. The geological proile was varying from soft soils with boulders
to rock. In total, 13 shafts were constructed with a diameter from 3.5–6.5
m, with the level of grouting pipes at 10 m depth from ground surface and
water pressures of 8 m. For the 60 buildings, an allowable settlement of 10 mm
290 Ground improvement

Treated area No treatment

Surface

Shaft
3.8mm

13.7mm

Tunnel west Tunnel east

Figure 7.32 Cross section of the City Tunnel Leipzig project with settlement curve
showing the difference between nontreated and grouted zone.

was speciied. More than 30,000 m of drilling and 1,100 liquid levelling
points were installed. The settlement reduction can clearly been seen in
Figure 7.32. In the areas outside the soilfrac treatment, a settlement of 13.7
mm occurred wheareas in the treated areas only 3.8 mm were registered.
A recent case of compensation grouting for a TBM excavation was made
for a Metro Line in Rome. Fracture grouting substituted the initially con-
sidered jet grouting operation, as less impact in terms of occupied public
space and discharge of spoil was given and in addition an active control
of the building protection was possible. Two EPB-tunnel-boring machines
passed underneath masonry and concrete buildings with a mimimum cover
of less than 3 m. More than 6,000 m of directional drillings were made,
settlements introduced due to drilling were about 1.5 mm although the
distance to the foundation was only 1 m in certain cases. The measured
settlements were less than 5 mm, signiicantly below the limit of 10 mm,
see Figures 7.33 and 7.34.

7.5.5 Application limits

The large number of practical applications shows that soil fracturing can be
applied under very different conditions with technical and economic advan-
tages. As far as improving soft soils is concerned, the application limit is
reached in very soft and organic soils. If the method is to be applied in
waste disposal, it is necessary to carry out a detailed preliminary chemical
 Soilfracture grouting 291

investigation. Regarding the geometrical conditions, it should be pointed

out once again that the treatment mainly modiies the stress low in the soil.
The concept of the design should therefore be formulated in the sense of an
improved and controlled stress transmission and should not be reduced to
the limited dimension of the foundation elements. The assumption that the

Monitoring
Artificial magnetic field

ca. 5.5 m
8.0 m

1.5 m
TAM array

Tunnel

Figure 7.33 Cross section of the Metro Rome settlement protection by means of
soilfracturing.

Monitoring

Fill

Grouted area Silt,

TAMs sandy, clayey
Tunnel

Tunnel Gravel

25m

Figure 7.34 Transversal cross section of the Metro B1 in Rome with Soilfrac area.
292 Ground improvement

dimension of the treated soil body must permit the delection of the stress
occurring can be the simple guide principle.
If there is doubt, large-scale tests have always been useful. These can be
integrated into subsequent main works and thus do not incur substantial
additional costs.

7.6 TEST AND CONTROL METHODS

The objectives of tests in connection with hydraulic fracturing can consist

of obtaining information on the injection technology itself, evaluating the
quality of soil improvement, or in assessing the effect on the structures con-
cerned. Usually, the intention is to draw conclusions on the achievable lifting
rates, the associated time requirements, and the material quantities involved.

7.6.1 Field trials

Experience has shown that reported tests are of value only if their dimen-
sions correspond in every way to those of the subsequent application.
However, the execution of such tests should be conceived as part of the
real application rather than as a kind of isolated basic test. In such cases,
there is no need to assess whether the test area selected comprises rep-
resentative soil and load conditions. What is important is to install a
set of geotechnical instruments providing reliable information and to
record all relevant parameters. The documentation of such a test has to
allow the evaluation of all parameters in terms of their time sequences.
As far as geotechnical instruments are concerned, apart from inclinom-
eters and extensometers, all systems are suitable which are able to record
deformation and changes in stress in the soil, the requirement being that
the deformability of the measuring elements themselves and that of the
annular grout should be similar to that of the surrounding soil; in the
case of pressure gauges, this is not always the case with every product
(Figures 7.35 and 7.36).
Suficient experience is available for shallow foundations in a wide range
of soils, whereas less knowledge is documented for buildings founded on
piles. Therefore real-scale ield trials are fundamentally important when
compensation grouting works have to be designed for the treatment of deep
foundations. It is necessary to consider the geometry of the pile (diameter,
length, spacing between piles) and to understand the bearing behaviour
of the single pile (end-bearing of skin-friction type) and the interaction
between the piles. The applied loads have to be those encountered in the
project (Figures 7.37 and 7.38).
Major ield trials in real scale have been performed for timber piles and
bored piles (Figures 7.37 and 7.38).
 Soilfracture grouting 293

A
Plan view
HI
HI

Extensometer HI Horizontal inclinometer

Drilling Vertical inclinometer Earth pressure cell
shaft 1
Electronic water level

Figure 7.35 Layout of a large-scale ield test which was planned as part of the subsequent
settlement compensation measure.

7.6.2 Laboratory tests

The suspensions provided as a skin-forming mixture, and for the injection itself
can vary considerably with respect to the number and type of components used
and their composition. In view of the different reactions of similar bonding
agents with different production origins, it is absolutely essential to carry out
basic tests at the start of each project. The effect of additives can especially be
inluenced by the local properties of cement and iller materials. The tests are
carried out with the objective of determining the low properties, the bleed-
ing of water and the setting behaviour as well as the stability of suspensions.
294 Ground improvement

Cross section A-A 2 mm

O 0.5 mm

970 kN 350 kN
350 kN
4 mm
4 mm
HI + 10.5 bar Shaft 1
3.2 mm + 0.2 bar 3.5 mm
+ 0.2 bar
HI
max. + 0.1 bar
+ 0.1 bar Vertical inclinometer
4 mm
Extensometer
Tunnel Electronic water level
Horizontal inclinometer
Earth pressure cell

Figure 7.36 Test results which show the suitablility of the measure for specifically
lifting individual foundations and the negligible influence on the tunnel
shell, respectively.

ca. 50 kPa

20 mm 30 mm 23 mm

Surface +/– 0.0

Fill

Piles
600 mm diameter

Clayey silt Electronic water level

Pressure cells
Precise levelling

–23.0 m

Gravel
Sleeve pipes

Figure 7.37 Cross section of the ield trial performed to prove that piled foundations
can be lifted with soil fracturing in a controlled manner.
 Soilfracture grouting 295

Diaphragm wall
Piles
600 mm diameter

A A

Electronic water level Pressure cell

Precise levelling Extenso-, inclinometer

Figure 7.38 Plan view of the ield trial performed with the monitoring system.

Although it is important to determinate special technical characteristics, the

most important premise remains the processibility of the suspensions using the
equipment available on site. While it is a common practice to vary the suspen-
sion composition in the course of the staged execution of the soil fracturing
work, it is advisable to set up a site laboratory. The objectives of suitability tests
can be the production of ‘soft mixtures’ to achieve low-end strength values of
granular soils as well as the production of ‘harder mixtures’ with the release of
small amounts of water for the purpose of improving soft soils.

7.6.3 Monitoring technology

The measuring technology used must allow the measures applied in the soil
and their effects on the soil to be very clearly associated with the structures
concerned. It is important to ensure that any blurred measured value does
not feign any movement tendencies which could lead to premature changes
in the works programme. Therefore, in case of doubt, reference arrange-
ments have to be used under controlled conditions. Such calibrations under
realistic conditions can concern all types of electronic and visual settle-
ment-measuring systems such as water level systems, automatic levelling
instruments and theodolites, rotary lasers, precision levelling instruments,
and special types such as loor-level measuring instruments, inclination
instruments, and crackmeter devices.

7.7 ENVIRONMENTAL AND CONTRACTUAL ASPECTS

Although hydraulic fracturing mainly manifests itself in the form of move-

ments in the soil and on structures, it is its ability to control the modiication
296 Ground improvement

of the soil and of the respective condition of the structure on which atten-
tion is focused. When agreements are drafted in connection with complex
structural methods, it is important to clarify at an early stage ownership
conditions and the interests of those who may be affected. The chemical
environmental compatibility of the individual injection components has to
be proved. In cases of doubt, additional tests of the actually used combina-
tion of materials and products have to be carried out.
Agreements regarding the realisation of projects involving hydraulic frac-
turing are concluded on the basis of mutually agreed projects. As some of
the effects of the measures applied can often not be speciied at that par-
ticular point in time, the agreement must permit the consistent use of the
observation method. The type of reaction to possible scenarios in the indi-
vidual construction phases has to be planned and contractually permitted,
of course with the intention of safeguarding the rights of third persons and
the economic execution of the project in the interest of solving an existing
problem. While taking into account the measures associated with strategies
against undesirable effects, it should be noted that a performance schedule
can list additional measures for limiting or warding-off injection effects.
Equally, the acceptance of crack formations and arching taking place in
the foot path region while a building is being lifted can be included in the
working agreement as part of the overriding project objective.

REFERENCES

Bernatzik, W. (1951). ‘Anheben des Kraftwerkes Hessigheim am Neckar mit Hilfe

von Zementunterpressungen’, Der Bauingenieur 26(4).
Brandl, H. (1981). ‘Stabilization of excessively settling bridge piers’, Florence:
Proceedings of the Xth ICSMFE, pp. 329–336.
Chambosse, G. and Otterbein, R. (2001). ‘Central Station Antwerp Compensation
grouting under high loaded foundations’, Istanbul: Proceedings of the XVth
ICSMFE.
Droof, E. R., Tavares, P. D., and Forbes, J. (1995). ‘Soil fracture grouting to remedi-
ate settlement due to soft ground tunnelling’, San Francisco: Rapid Excavation
and Tunnelling Conference.
Falk, E. (1997). ‘Underground works in urban environment’, Hamburg: Proceedings
of the XIVth ICSMFE.
Falk, E. and Schweiger, H. F. (1998). ‘Shallow Tunneling in Urban Environment –
Different Ways of Controlling Settlements’, Felsbau 16(4).
Haimoni, A. M. and Wright, R. H. (1999). ‘Protection of Big Ben using compensation
grouting’, Ground Engineering 32(8):33–37.
Harris, D. I. (2001) ‘Protective measures’, London: Response of buildings to excava-
tion–induced ground movements, CIRIA Conference.
Jakobs, M., Otterbein, R., and Dekker, H. (2001). ‘Erfahrungen beim Einsatz
der Druckschlauchwaage zur Höhenüberwachung setzungsempindlicher
Bauwerke’, Zeitschrift Bauingenieur (76).
 Soilfracture grouting 297

Kaalberg, F.J., Essler, R.D., and Kleinlugtenblet, R. (2011). ‘Compensation grout-

ing of piled foundations to mitigate tunnelling settlements’, Proceedings of
7th International Symposium on Geotechnical Aspects of Underground
Construction in Soft Ground.
Kudella, P. (1994). Mechanismen der Bodenverdrängung beim Einpressen von
Fluiden zur Baugrubenverfestigung, Karlsruhe. Veröffentlichungen des
Institutes für Bodenmechanik und Felsmechanik der Universität Fridericiana.
Kummerer, C. and Passlick, T. (2008). ‘Kompensationsinjektionen als besondere
Schutzmaßnahmen von Bauwerken bei Tunnelunterfahrungen’. Tagungsband
zum 23. Ch. Veder Kolloquium, Institute for Soil Mechanics and Geotechnical
Engineerging, Graz University of Technology, 131–144.
Mair, R. J., Harris, D. I., Love, J. P., Blakey, D., and Kettle, C. (1994). ‘Compensation
grouting to limit settlements during tunnelling at Waterloo Station, London.’
Tunnelling 94:279–300.
Raabe, E. W. and Esters, K. (1993). ‘Soilfracturing techniques for terminating settle-
ments and restoring levels of building and structures.’ In: Ground Improvement,
Chapman & Hall, Glasgow.
Sagaseta, C., Sánchez-Alciturri, J. M., González, C., López, A., Gómez, J., and
Pina, R. (1999). ‘Soil deformations due to the excavation of two parallel cav-
erns.’ Rotterdam: Proceedings of the Twelfth European Conference on Soil
Mechanics and Geotechnical Engineering.
Samol, H. and Priebe, H. (1985). ‘Soilfrac – ein Injektionsverfahren zur
Bodenverbesserung’, San Franciso: Proceedings of the XIth ICSMFE.
Schweiger, H. F., Falk, E. (1998). ‘Reduction of settlements by compensation
grouting – Numerical studies and experience from Lisbon underground’,
Brazil: Proceedings of the World Tunnel Congress 1998 on Tunnels and
Metropolises, Sáo Paulo.
Schweiger, H. F., Kummerer, C. Otterbein, R. (2004). ‘Numerical modelling of settle-
ment compensation by means of fracture grouting’, Soil and Foundation, J. of
Japanese Geotechnical Society, 11(1): 71–86.
Sciotti, A., Desideri, A., Saggio, A. Kummerer, C. (2011). ‘Mitigation of the effects
induced by shallow tunneling in urban environment. The use of “compen-
sation grouting” in the Underground Line B1 work’. Proceedings of the
Seventh International Conference on Geotechnical Aspects of Underground
Construction in Soft Ground, Rome (in print).
Van der Stoel, A. E. C., Haasnoot, J. K., Essler, R. D. (2001). ‘Feasibility of
Compensation Grouting of Timber Pile Foundations to Mitigate TBM
Settlements, Full Scale Trial North/South Line Amsterdam’, London: CIRIA
Conference.
Chapter 8

Compaction grouting
James Hussin

CONTENTS

8.1 Introduction ................................................................................. 300

8.2 History ......................................................................................... 302
8.3 Applications.................................................................................. 304
8.4 Applicable soil types ..................................................................... 306
8.5 LMG mix design .......................................................................... 306
8.6 Design of the LMG programme.................................................... 308
8.6.1 Subsurface conditions ........................................................ 308
8.6.2 Constraints ........................................................................ 308
8.6.3 Requirements..................................................................... 309
8.6.4 Criteria .............................................................................. 309
8.6.5 Layout................................................................................ 310
8.6.5.1 Spacing ................................................................. 310
8.6.5.2 Injection sequence ................................................ 310
8.6.6 Procedures ......................................................................... 311
8.6.6.1 Injection pressure ................................................. 311
8.6.6.2 Injection rate ........................................................ 311
8.6.6.3 Injection volume ................................................... 311
8.6.6.4 Heave ................................................................... 312
8.7 LMG programme procedures ....................................................... 312
8.7.1 Layout and sequence.......................................................... 312
8.7.2 Pipe installation ................................................................. 313
8.7.3 Grout injection................................................................... 313
8.8 Equipment .................................................................................... 314
8.8.1 Batchers and mixers........................................................... 314
8.8.2 Pumps ................................................................................ 315
8.8.3 Injection pipe installation equipment ................................. 315
8.8.3.1 Hand equipment................................................... 316
8.8.3.2 Tracked and wheeled drill rigs.............................. 316
8.8.4 Miscellaneous .................................................................... 316
8.8.4.1 Extraction of pipe................................................. 317

299
300 Ground improvement

8.8.4.2 Hoses.................................................................... 317

8.8.4.3 Pressure gauges with gauge savers ........................ 317
8.9 QC & QA..................................................................................... 317
8.9.1 LMG characteristics .......................................................... 319
8.9.2 Test programme................................................................. 319
8.9.3 Soil improvement testing.................................................... 320
8.9.4 Quality control during injection ........................................ 320
8.9.5 Grout injection pressure..................................................... 320
8.9.6 Grout injection rate............................................................ 320
8.9.7 Volume of grout injected.................................................... 320
8.9.8 Heave................................................................................. 321
8.9.9 Instrumentation, electronic monitoring, and computer
data acquisition.................................................................. 321
8.10 Case histories................................................................................ 321
8.10.1 Harlem Hospital, New York, New York............................ 321
8.10.2 Interstate 5 at Hasley Canyon, California.......................... 322
8.10.3 Tunnelling beneath Highway 85, Denver, Colorado .......... 326
References ............................................................................................. 328

8.1 INTRODUCTION

Compaction grouting involves the subsurface injection of a stiff mortar

grout. Since its inception in the mid-twentieth century, its use has evolved
to address many subsurface problems. This has produced discussion within
the industry as to the proper deinition of compaction grouting, compo-
nents and characteristics of the grout, and the proper procedures that
should be followed.
Injection of thick mortar grout was initially performed in the 1950s to
ill relatively small voids beneath structures (Warner, 1982). It was soon
discovered that after a void was illed, additional pumping resulted in ‘jack-
ing’ of the overlying structure, which made the process a valuable relevel-
ling tool. It wasn’t until the late 1960s that the side effect of soil compaction
was identiied (Graf, 1969). This is when the term ‘compaction grouting’
was coined.
The low mobility grout became known as ‘compaction grout’ and the
process of injecting it, ‘compaction grouting’. However, as when it was
originally developed, the grout’s low mobility characteristic made it attrac-
tive for other applications such as illing of subsurface voids, either man-
made (such as mines) or naturally occurring (such as karst conditions), or
for constructing high modulus columns to reinforce soft wet soils that were
not compacted by the process. These applications often did not involve soil
compaction, and some in the industry felt the term ‘compaction grouting’
was being improperly applied. To help resolve this controversy, the term
 Compaction grouting 301

‘low mobility grout’ (LMG) was proposed in the late 1990s (Byle, 1997) to
describe the grout and the broad process of using LMG. Compaction grout-
ing would be a subcategory of LMG.
In 2010, the Grouting Committee of the Geo-Institute of the American
Society of Civil Engineers (G-I ASCE) published the ‘Compaction Grouting
Consensus Guide’ (ASCE/G-I 53-10). In this guide, compaction grouting is
deined as follows:

Compaction Grouting is a ground improvement technique that improves

the strength and/or stiffness of the ground by slow and controlled injec-
tion of a low mobility grout. The soil is displaced and compacted as the
grout mass expands. Provided that the injection process progresses in a
controlled fashion, the grout material remains in a growing mass within
the ground and does not permeate or fracture the soil. This behavior
enables consistent densiication around the expanding grout mass,
resulting in stiff inclusions of grout surrounded by soil of increased den-
sity. The process can be applied equally well above and below the water
table. It is usually applied to loose ills and loose native soils that have
suficient drainage to prevent buildup of excess pore pressure.

Although change is slow, it appears that the industry as a whole is in

the process of adopting the term ‘low mobility grout (LMG)’ to describe
the grout itself and its broad use, and the term ‘compaction grouting’ only
when the goal is to compact soils (Figure 8.1). Since this transition is still
in progress, this chapter shall discuss the general use of LMG, both when

Figure 8.1 Compaction grouting to densify soil beneath existing foundation. (Courtesy
of Hayward Baker Inc.)
302 Ground improvement

used to compact soils and for other applications. The term ‘low mobility
grout (LMG)’ shall be used for the grout.

8.2 HISTORY

Compaction grouting is the only major grouting method originating in

the United States. The following timeline presents a general history of the
development of the LMG industry:

Early 1950s A small contractor in California irst pumped a stiff mortar-

type grout to ill voids beneath distressed structures and to
relevel settled foundations and slabs.
Late 1960s Compaction of the soils surrounding injected mortar grout
was identiied and the term ‘compaction grouting’ was
irst used.
Late 1970s Compaction grouting was irst used to compact soils loos-
ened during soft ground tunnelling for the Baltimore sub-
way to prevent settlement of overlying structures (Figure
8.2). The process was so successful that it was also used for
other tunnelling projects, including the subways for Boston,
Seattle, and Los Angeles.
Early 1980s Compaction grouting was used for large-scale precon-
struction soil improvement to densify loose sands beneath
a planned power plant in Jacksonville, Florida (together
with dynamic compaction), and beneath many structures
for the planned nuclear submarine base in Kings Bay,
Georgia (together with dynamic compaction and vibro-
replacement) (Hussin, 1987) (Figure 8.3).

Protected
building
Settlement foundations
without
grouting

Compaction
grout bulb
Compaction grout bulb

Tunnel
Tunnel

Figure 8.2 Compaction grouting to prevent settlement of overlying foundations during

soft ground tunnelling. (Courtesy of Hayward Baker Inc.)
 Compaction grouting 303

Mid-1980s Prior to this time, slurry grout (Portland cement and water)
was used to stabilise sinkhole conditions in karst regions.
However, since the grout was very luid and the limestone
contained many cracks and crevasses, the required grout
quantity was very dificult to predict, resulting in frequent
large cost overruns. A specialty contractor in Florida
(Henry, 1986) realised that LMG offered the beneit of
illing the larger voids and compacting loosened soil while
at the same time, its stiff nature would limit grout takes in
the thin cracks of the limestone formation (Figure 8.4). The
process was very successful and essentially replaced slurry
grouting to stabilise sinkholes, compact resulting loos-
ened sands, and relevel overlying settled structures. Soon
after, the technique saw its irst use to treat sites in sink-
hole prone areas prior to construction to reduce the poten-
tial for future sinkholes. This application has since been
adopted in karst geology regions throughout the United
States. By the late 1980s the technique began to be applied
in Europe, notably the UK, mainly for ground improve-
ment and foundation remedial works (see Crockford and
Bell, 1996).
Circa 1990 Compaction grouting was exported to Japan where it
began to gain acceptance to raise structures and was later
used to relevel structures that settled as a result of the
Kobe earthquake (1995).

Figure 8.3 Compaction grouting combined with vibro-replacement and dynamic com-
paction beneath planned structures at Georgia submarine base. (Courtesy of
Hayward Baker Inc.)
304 Ground improvement

Figure 8.4 LMG to stabilise large sinkhole at a central Florida mining facility. (Courtesy
of Hayward Baker Inc.)

The news of the applications and successes spread and LMG is now used in
many parts of the world. The term ‘compaction grouting’ is still used outside
the United States to mean the whole process used in the full range of applica-
tions including ground densiication, relevelling, or compensation using LMG.

8.3 APPLICATIONS

The low mobility characteristic of LMG makes it suitable for modifying

subsurface conditions both beneath existing structures and prior to con-
struction. Common applications include

r Preconstruction soil improvement to permit shallow foundations

r Compaction of loose ills
r Creation of grout columns and illing voids within loose or deteriorat-
ing natural soil conditions (i.e., organic degradation, etc.) or voided ills
r Compaction of loose soils resulting from adjacent excavation activ-
ity, tunnelling (Figure 8.5d), sinkhole activity (Figure 8.5b), improper
dewatering, broken utility lines, and so forth
r Compaction to increase bearing capacity or reduce potential settle-
ment beneath existing foundations when modiications to the existing
structure increase the foundation loading (Figure 8.5c)
r Compaction of deep loose zone (static or dynamic settlement potential)
 (a) (b) (c)

(d) (e) (f)

Compaction grouting
Figure 8.5 Low mobility grout applications. (Courtesy of Hayward Baker Inc.)

305
306 Ground improvement

r Improvement of ground between pinnacled limestone to avoid deep

piling within slots (Figure 8.5e)
r Injection of grout beneath settled structures to heave and relevel the
structures (Figure 8.5a and f)
r Increasing lateral support for existing deep foundations (Figure 8.5c)
r Injection of grout to compensate for ground movements arising from
underlying tunnelling operations or near deep excavations

8.4 APPLICABLE SOIL TYPES

LMG has been applied in almost all soil types. The process can be con-
trolled to construct a column of mortar grout which acts as a reinforcing
element in the soil. The ability of the process to compact the surrounding
soils (compaction grouting) is inluenced by the soil’s properties.
The process of compacting soil constricts the void spaces, requiring any
water within the void to exit. Therefore, compaction grouting is most effec-
tive in soils which have a high permeability and/or low degree of satura-
tion. Loose, highly permeable granular soils are best suited for compaction
grouting. As the ines content increases (particularly clay content), the soil
permeability decreases along with the effectiveness of compaction grout-
ing. Sands with less than 10% silt and no clay compact well above and
below the water table. Some limited compaction has been accomplished in
nonsaturated, nonplastic, ine-grained soils with a very slow grout injec-
tion rate and a carefully designed grout mix (detailed later in this chapter).
Collapsible soils have also been successfully treated with compaction grout-
ing. Soils best suited for compaction grouting are presented in Figure 8.6.
LMG is particularly effective as a load transfer element in noncom-
pactable soil layers where the layer thickness is only several feet thick. In
this situation, a compaction grout column with a minimum diameter equal
to half of the layer thickness can effectively support the load that would
otherwise induce stress in the layer. As previously stated, LMG has been
effectively used to ill subsurface voids and to build columns in abandoned
mines to provide roof support (Figure 8.7).

8.5 LMG MIX DESIGN

The required characteristics of LMG vary depending on the application.

LMG used for compaction grouting is generally most restrictive and will
irst be discussed. Afterwards, other applications for which the require-
ments can be eased will be discussed.
LMG has what may appear to be the conlicting requirements of being
pumpable yet immobile. The grout is most effective if it has enough ines to
 Compaction grouting 307

Gravel Sand Silt & Clay

100

Reinforcable
Unsaturated with drainage
Percent passing

50

30

Saturated

10

0
10 5 2 1.0 0.5 0.2 0.1 0.05 0.02 0.01 0.005
Particle size - mm

Figure 8.6 Range of soil gradation curves best suited for compaction grouting. (Courtesy
of Hayward Baker Inc.)

maintain its pumpability but be suficiently permeable to give up its water

once it exits the end of the pipe. The gradation of the grout’s aggregate is
most inluential in achieving this property. A commonly used gradation is
a uniformly graded aggregate, including gravel to silt size material with the
100% iner than the 5–20 mm range and the 0% iner than the range of
0.001–0.03 mm. This gradation will generally maintain the grout’s internal
friction and permeability, resulting in an immobile grout once it leaves the
end of the pipe in the ground.
Care should be used when adding bentonite clay to LMG. Although benton-
ite makes the grout more pumpable, the percentage should be kept low (<1%
by total dry weight of solids) to avoid excessively reducing the grout’s perme-
ability and possible fracturing within the soil. Bentonite has been successfully
used when compacting low ines content sands and injecting the grout at a
very low rate (~2 cubic feet or 0.06 cubic metres per minute.) The use of ben-
tonite in LMG is less of an issue in void-illing and karst applications.
The amount of water included in the mix should be the minimal amount
required for the grout to be pumpable. This generally results in a very stiff mix.
For compaction grouting applications, the primary goal is generally com-
paction of the surrounding soils, and the grout only needs to be as strong as the
soil. Therefore, cement is not a requirement. However, cement is often used to
308 Ground improvement

Figure 8.7 LMG used to ill a mine tunnel. (Courtesy of Hayward Baker Inc.)

provide the ines necessary to make the grout pumpable and is required when
the design requires the grout to carry load or span a void as in the case of cap
grouting in karst applications. Other materials used as ines in LMG mixes
include lyash and silt (when natural silty ine sand is used in the mix).

8.6 DESIGN OF THE LMG PROGRAMME

Several important steps are involved in the design of a LMG programme.

8.6.1 Subsurface conditions

It is critical to understand the subsurface conditions in order to design the
LMG programme to achieve the desired results. The stress conditions in the
ground prior to grouting will inluence the design quantities and injection
pressure. Identifying variations in soil density with depth allows the pro-
gramme to target loose strata and avoid wasted effort in dense strata. Soft
strata may only require a certain diameter grout column for reinforcement.
The rate of injection may be able to be varied based on the permeability
of different strata. The quantity of grout to be injected will depend on any
subsurface voids or hard inclusions.

8.6.2 Constraints
The above- and below-ground constraints must be understood prior to
designing the grouting programme. Some areas of the site may not be acces-
sible or may require limited access equipment or hand-held equipment.
Both above- and below-ground utility locations must be carefully identi-
ied. Subsurface structures or adjacent retaining walls must be identiied to
avoid damage during the high-pressure injection of the grout.
 Compaction grouting 309

8.6.3 Requirements
The requirements of the programme are important to understand before
the LMG programme can be designed. The requirements could involve
densiication of the soils, reinforcement with grout columns, relevelling of
foundations or compensation for loss ground due to tunnelling.
If densiication is the target, the minimum required post-treatment
density must be determined. Relevelling of an existing structure could be
necessary. In the scenario of compaction grouting above soft ground tun-
nelling to avoid excessive settlement of overlying structures, the maximum
allowable settlement must be deined.

8.6.4 Criteria
The criteria that deine success or failure must be deined and understood
by all involved parties. Common acceptance criteria include injected grout
volume, injected grout pressure, inal elevation of lifted structures and test
results (SPT, CPT, etc.) of soil between injection locations. Often, the same
test is performed both before and after grouting to allow an accurate deter-
mination of the improvement (see Figure 8.8).

308

306 Pre-grout N values

304
Post-grout N values
Elevation (ft) [1 ft = 0.3 m]

302

300

298

296

294

292

290
0 10 20 30 40 50 60
N value
(blows/ft = blows/0.3 m)

Figure 8.8 Sample SPT results before and after compaction grouting. (Courtesy of
Hayward Baker Inc.)
310 Ground improvement

8.6.5 Layout
The design of the grout injection pipe layout is based on the subsurface condi-
tions, site constraints, programme requirements, and the criteria to be achieved.

8.6.5.1 Spacing
The spacing of the pipes in plan depends primarily on the soil type and
the criteria to be achieved. Spacing generally varies between 3–15 feet
(0.9–4.6 m) on centre. These are extreme examples. The spacing is on the
lower end of the range for shallow treatment (low overburden or overly-
ing structural load) and in ine-grained soils. Spacing is on the higher end
of the range for deeper treatment (greater overburden or heavier overly-
ing structural load), in granular soils, or when limited improvement is
required. Most compaction grouting is performed at a spacing of 5–7 feet
(1.5–2.1 m) (see Figure 8.9).

8.6.5.2 Injection sequence

The sequence by which the injection is performed is important to maximise
the improvement achieved. For a given layout pattern, the best results are
achieved by injecting at every other location irst (primary locations) followed

Plan view - Compaction grouting layout

5.66 ft
P (1.72 m) P P P P P
Typical
8 ft
(2.4 m) S S S S S
Typical
P P P P P P

S S S S S

P P P P P P

S S S S S

P P P P P P

P = Primary location Post treatment test location

S = Secondary location

Figure 8.9 Sample compaction grouting layout plan. (Courtesy of Hayward Baker Inc.)
 Compaction grouting 311

by the skipped locations (secondary locations), see Figure 8.9. The primary
injections provide coninement of the soils during the secondary grouting.
The vertical sequence at a speciic injection location can also be varied.
LMG may begin at the bottom of the treatment zone and proceed upward
(upstage) or at the top of the treatment zone and proceed downward (down-
stage). The upstage process involves installation of the grout pipe to the
bottom of the treatment zone and then slowly extracting the pipe as the
LMG is injected. The upstage process is generally easier, quicker, and less
costly. The downstage process involves installation of the grout pipe to the
top of the treatment zone, injecting the LMG at that depth, waiting for the
grout to set, drilling through the grout, and then pumping grout beneath
the previously injected grout. The process is repeated until grout has been
injected into the full treatment depth. This process is more effective when
lifting structures from a shallow depth and shallow weak soils.

8.6.6 Procedures
8.6.6.1 Injection pressure
Injection of LMG requires relatively high pressures, possibly in excess of
1,000 psi (6.99 MPa). It may require 100–200 psi (0.7–1.4 MPa) just to
pump the stiff LMG through the grout hose and injection pipe before the
pipe is inserted into the ground. Once the pipe is inserted into the ground,
the additional pressure required to displace and compact the surrounding
soils increases with depth and with soil density. The pump injection pres-
sure will typically increase as the injection continues at a particular loca-
tion due to the increase in density of the surrounding soils.

8.6.6.2 Injection rate

The maximum effective injection rate for LMG depends on the permeabil-
ity of the soils being treated. As the ines content increases, the injection
rate should be lowered. A clay content of 1%–2% can greatly reduce the
permeability of granular soil and lower the maximum effective injection
rate. In low permeability soils or low coninement situations (shallow depth
and no overlying structures) an injection rate of 0.5–1 ft3 (14.2–28.3 litres)
per minute is common. In well-draining soils, dry soils, or soils at depths,
4–12 ft3 (113–340 litres) per minute is appropriate. The injection rate is not
limited in void-illing applications.

8.6.6.3 Injection volume

The injection volume depends on the density of the soil and the amount of
improvement required. The volume reduction from a granular soil’s mini-
mum density to its maximum density may be 15%–20%. However, typical
312 Ground improvement

injection volumes are in the range of 8%–12% of the soil to be compacted.

The volume may require adjustment as the programme progresses.

8.6.6.4 Heave
Heave typically occurs when the resistance to grout displacement is less
above the grout than it is laterally. It can also be an indication that fractur-
ing has occurred. Heave is a limiting factor in that when it occurs, little
additional compaction is occurring in the surrounding soils. Heave can
also damage overlying structures. However, if the objective is to restore
the levels of an overlying structure which has previously experienced settle-
ment, some heave may be desirable to relevel the structure.
Heave can occur when grouting at shallow depths. It is dificult to den-
sify soils at depths shallower than 10 feet (3 m) because the limited over-
burden pressure is less than the lateral pressure of the soils. Heave can also
occur during deeper grouting when the surrounding soils have been com-
pacted such that they resist further displacement. Once heave is detected,
grouting is generally discontinued at that depth. Compaction of overlying
loose zones can still be achieved.
The heave criteria should be established before the grouting programme
begins. It is important to realise that heave is often cumulative. Therefore,
the maximum heave criteria at any particular depth should be less than the
maximum allowable for the overlying structure to allow continued grout-
ing in soils above the depth where the heave criterion is irst observed. For
example, if 0.5 inch (12.7 mm) is the maximum total allowable heave, a rea-
sonable refusal criterion for any particular depth might be 0.1 inch (2.5 mm).

8.7 LMG PROGRAMME PROCEDURES

The LMG programme procedures are established prior to the programme

beginning and assist in making sure that the parameters established during
the design are followed. A close working relationship between the special-
ist contractor and the client’s ield inspector is beneicial in achieving the
programme’s goals. Monitoring of the work and surrounding structures is
an integral part of the procedures.

8.7.1 Layout and sequence

The plan layout of the injection locations and sequence in which they are
performed should be carefully followed. Sequence includes the order that
the locations are performed, including if a primary/secondary sequence is
to be followed. Also, the sequence includes either an upstage or downstage
procedure at each location.
 Compaction grouting 313

8.7.2 Pipe installation

The main concern when installing the grout pipe is that the surrounding
soil is in tight contact with the exterior of the pipe. This is important to
avoid grout travelling up the annulus between the soil and pipe exterior as
well as to provide resistance against the pipe jacking up out of the ground
during pumping. Installation methods include driving, lushing, and drill-
ing. Problems associated with driving the pipe include encountering refusal
prior to reaching the target depth, damaging the pipe during driving,
and negatively affecting existing structures due to the driving vibrations.
Flushing can be external or internal. External lushing should only be used
with care since it can cause an open annulus. It has been used success-
fully in granular soils in limited amounts since the granular soils collapse
around the pipe at the end of lushing. Internal lushing is generally accept-
able. Drilling is the most common installation procedure since it avoids
these problems. When lushing is used in conjunction with drilling, these
comments concerning lushing apply. Upstage and downstage sequencing is
discussed above in Section 8.6.5.2. When performing downstage grouting,
it is common to have to wait 8 to 10 hours for a stage to achieve initial set
before advancing the pipe through it to grout the underlying stage.
During pipe installation the location and angle of insertion should be
carefully monitored and documented so that the grouting is performed in
the planned location and no locations are missed or repeated. The dificulty
of installation should also be monitored and documented as an indication
of the soil density with depth prior to grouting.

8.7.3 Grout injection

The drilling and grouting operations can be performed by the same crew
and equipment, completing both at each location before proceeding to
the subsequent location, or they can be performed by separate crews; one
installing the pipes and another following behind performing the grouting
and pipe removal.
Pumping water through the grout hoses prior to pumping grout prevents
the loss of water from the irst stroke of grout mix as it is pumped through
the hose, avoiding possible plugging. It is also important that all hose and
pipe connections are water-tight to avoid water loss and possible grout
plugging at these locations. Before connecting the grout hose to the top of
the injection pipe, grout should be pumped to ill the hose.
Often the volume of grout is determined by counting the number of pump
strokes. Prior to beginning the programme, the grout pump should be cali-
brated. One simple method of calibration consists of counting the number
of strokes required to ill a container of ixed volume, such as a 55-gallon
(208-litre) drum.
314 Ground improvement

During grout injection the grout pressure and volume should be moni-
tored and recorded versus depth. The pressures in the hose should be moni-
tored and recorded both near the pump and near the top of the injection
pipe (requiring pressure gauges at these locations). The pressure at the
top of the pipe is a close indication of the injection pressure into the soil.
A comparison of the two gauges reveals if high pressures or grout refusal is
a result of grout plugs within the line.
The three parameters that often are speciied as controlling when a grout-
ing stage is complete are grout injection pressure, grout volume, and ground
heave. The reasoning behind specifying a maximum pressure is that when
achieved, the soils surrounding that injection stage have been suficiently com-
pacted. A maximum volume is speciied when the design requires a grout
column of a deined diameter to carry a load or possibly at primary locations
to provide coninement of the soils during the subsequent secondary location
grouting. A maximum heave is speciied to avoid excessive grouting without
additional compaction and to avoid damage to overlying structures. Grouting
is often continued at a stage until one or the other of these is irst observed.

8.8 EQUIPMENT

Proper equipment is integral to the success of a LMG programme and par-

ticularly to a compaction grouting programme.

8.8.1 Batchers and mixers

The batcher and mixer requirements vary depending on the daily grout
volume required and the grout mix requirements, both of which depend
on the application. When LMG is used in void-illing or karst applications,
the internal friction of the grout is not as critical and often is more easily
mixed (i.e., may simply contain a ine sand aggregate and, possibly, benton-
ite). These mixes may be provided by a ready-mix plant and mixed during
transport in the revolving drum of a common concrete truck (Figure 8.10).
However, when performing compaction grouting, the internal friction is
often required and this mix does not mix well in a truck. Also, the lower
volumes typical of compaction grouting do not suit ready-mix plants and
concrete trucks because of the signiicant time each load might be delayed
in the truck while waiting to be pumped.
For compaction grouting applications, on-site batching is typical. Pugmill
mixers usually are slow and labor intensive (adding materials by hand) and
are only used when small quantities are expected (less than 5 cubic yards [3.8
cubic metres] per shift). For projects requiring larger quantities, a continuous
mix batch plant is often used. These batch plants can vary in size and capacity
and are often mounted on a truck. Typically, the components are metered onto
 Compaction grouting 315

Figure 8.10 Concrete truck delivers LMG to grout site. (Courtesy of Hayward Baker
Inc.)

Figure 8.11 Examples of on-site LMG batch plants. (Courtesy of Hayward Baker Inc.)

a belt which travels beneath the material storage bins. The materials are then
mixed with water in a continuous screw auger which then feeds the grout into
the pump. These batch plants can produce as much as 50 cubic yards of grout
(38 cubic metres) per hour. Sample batch plants are depicted in Figure 8.11.

8.8.2 Pumps
LMG is typically pumped with modiied concrete pumps. These are piston
pumps which pump a deined volume per stroke (the volume of the piston).
The typical concrete pump generally requires modiication of the piston
diameter and control mechanisms to be capable of pumping at the high pres-
sures (1,000 psi or 6.9 MPa or greater) and slow rates of less than 2 cubic
feet (57 litres) per minute required to perform compaction grouting.

8.8.3 Injection pipe installation equipment

The method of installing the grout injection pipe can vary depending on
site constraints and the application.
316 Ground improvement

Figure 8.12 Hand equipment to install LMG injection pipe. (Courtesy of Hayward Baker
Inc.)

8.8.3.1 Hand equipment

Many LMG projects are in tight environments with injection location barely
accessible to workers alone (Figure 8.12). In these situations, the pipes can
be installed with hand-held equipment. These can consist of hand-held
rotary percussion drills, or small hydraulic or pneumatic driving equipment.

8.8.3.2 Tracked and wheeled drill rigs

When more room is available, tracked or wheeled equipment may be
suitable. The size of this equipment varies from very small that can it
through a standard doorway, to very large and crawler-crane mounted,
see Figure 8.13. Generally, equipment with the longest mast (single stroke)
is most productive in that it can install longer sections of pipe at a time.
Augering a hole and then inserting the grout pipe is generally not desirable
since it does not result in a tight it between the pipe and the soil.

8.8.4 Miscellaneous
Several miscellaneous items are also important to consider.
 Compaction grouting 317

Figure 8.13 Small-wheeled and large-tracked drills to install LMG injection pipe.
(Courtesy of Hayward Baker Inc.)

8.8.4.1 Extraction of pipe

Extraction of the pipe during grouting can be performed by the same
equipment that inserted it (drill or crane) or could be performed as a sepa-
rate operation. When handled separately, purpose-built hydraulic rams are
often used (Figure 8.14).

8.8.4.2 Hoses
High-pressure hoses or steel pipes with lush joints and a minimum diam-
eter of 2 inches (51 mm) are required to avoid excessive resistance to pump-
ing the grout through the lines. The ittings (couplings) between sections
of hose or pipe should be tight to avoid water leaks which could lead to
blockages in the lines.

8.8.4.3 Pressure gauges with gauge savers

Pressure gauges are necessary to monitor the grout injection pressure. As
noted in Section 8.7.3, gauges should be located both near the pump and
near the top of the injection pipe (Figure 8.15).

8.9 QC & QA

A properly speciied quality control (QC) and quality assurance (QA) pro-
gramme is essential to the successful performance of a LMG programme.
318 Ground improvement

Figure 8.14 Hydraulic ram to extract LMG injection pipe. (Courtesy of Hayward Baker
Inc.)

Figure 8.15 Pressure gauge and gauge saver near top of grout injection pipe. (Courtesy
of Hayward Baker Inc.)
 Compaction grouting 319

The programme should be designed to specify, monitor, and measure the

parameters necessary for the speciic application.

8.9.1 LMG characteristics

Depending on the application, speciic aspects of the LMG should be mon-
itored and documented. A laboratory mix design is typically performed
prior to the grouting work to ensure that the grout will achieve the required
parameters (strength, slump, etc.). The results of a previously performed
mix design programme could apply if the same components were used. If
the grout is batched on site, the ratio of the components should be accu-
rately measured and documented to assure a consistent product.
If the design requires minimum grout strength, then test specimens, typi-
cally cylinders, should be cast on a regular interval. This interval may be
based on time (e.g., twice a day) or grout volume (e.g., every 100 cubic yards
[76 cubic metres]). If grout slump is important, it is common for the slump
measurements to be performed when the UCS samples are cast (Figure 8.16).

8.9.2 Test programme

For many larger projects, a test programme is performed to demonstrate
that the planned materials and procedures will achieve the required results.
A test programme may be performed within or separate from the pro-
duction area. If densiication of soils is required (compaction grouting),
penetration testing of the target soils before and after treatment is often
performed. If the LMG is planned to reduce settlement potential in soils
which may not show signiicant improvement in penetration test values
(ine-grained soils), a load test may be performed of an individual grout
column or of a treated area.

Figure 8.16 Casting test cylinders and measuring LMG slump. (Courtesy of Hayward
Baker Inc.)
320 Ground improvement

8.9.3 Soil improvement testing

When soil improvement is required, a performance speciication is often
selected. Typically, the acceptance criterion requires that post-treatment
penetration testing of the soil midway between the grout injection loca-
tions achieves a minimum and/or average value. Common post-treatment
penetration testing could include Standard Penetration Tests (SPT), Cone
Penetrometer Tests (CPT), and Dilatometer Tests (DMT). The speciied
value should be based on an analysis of the required performance (bearing
capacity, settlement, liquefaction, etc.). An example of testing location and
results is depicted in Figures 8.8 and 8.9.

8.9.4 Quality control during injection

Several parameters should be monitored and recorded during grout injection.

8.9.5 Grout injection pressure

The grout pressure during injection should be measured in the grout hose
both near the pump and near the top of the grout injection pipe (as explained
in Section 8.7.3). The LMG will require a signiicant pumping pressure to
overcome the line friction and passive resistance of the soil at the bottom of
the injection pipe. An initial injection pressure in the range of 100–200 psi
(0.7–1.4 MPa) is common. The pressure required to inject additional grout
increases as the soil is displaced and compacted. Final injection pressures
may exceed 1,000 psi (6.9 MPa).

8.9.6 Grout injection rate

The grout injection rate should be monitored and recorded. The com-
mon method is by calibrating the pump piston (as mentioned in Section
8.7.3) and then counting the number of piston strokes during injection.
Controlling the rate of grout injection is important when performing com-
paction grouting. If the rate is too great, the soil pore water may not have
time to drain and result in an increase in pore water pressure. The pore
pressure increase may result in an increase in grout pumping pressure, giv-
ing a false indicator of soil improvement. The typical grout injection rate
is between 1–5 cubic feet (28–142 litres) per minute. A gradual increase in
injection pressure indicates controlled densiication. A sudden drop in pres-
sure indicates hydraulic fracturing.

8.9.7 Volume of grout injected

The grout volume injected should be monitored and recorded. Depending
on the initial density of the soil being treated, the grout injection volume
 Compaction grouting 321

typically required to compact the soil is between 5%–15% of the volume of

the soil being treated. The volume of the soil being treated extends halfway
to the adjacent injection locations. If signiicantly more grout is injected
without a signiicant increase in injection pressure, the location should be
explored for subsurface voids or utilities. When a primary/secondary inject-
ing sequencing is planned, the volume injected at the primary locations is
typically restricted to a maximum of 15% with the goal of even grout dis-
tribution. This is necessary to determine payment in unit price contracts.

8.9.8 Heave
During grout injection, the grout will displace the ground in the direction of
least resistance. In loose soils and when injecting at depth, the overburden
weight provides more resistance than the lateral passive resistance, result-
ing in the grout displacing the soil horizontally. As the soils become denser
and when injecting at depths shallower than about 10 feet (3 m), the path
of least resistance may be towards the ground surface. This may result in
heave of the ground surface and overlying structures. Little additional soil
compaction is occurring during heave. Many different instruments are used
to monitor when heave occurs and the magnitude of the heave. Examples
include crack monitors, tilt meters, plumb bobs, and spirit levels.

8.9.9 Instrumentation, electronic monitoring,

and computer data acquisition
Computers have become more capable of withstanding the vibrations and
dust associated with construction, allowing their incorporation into many
aspects of construction. Computers can be connected to instrumentation
that measures grout injection pressure, volume and depth. The computer
along with the instrumentation is referred to as a data acquisition (DAQ)
system. This is not only useful for documentation of the work, but can also
allow the operators to monitor their work and make adjustments as neces-
sary, since measurements are displayed on a screen in real time (Figure 8.17).

8.10 CASE HISTORIES

The following case histories present several applications for LMG.

8.10.1 Harlem Hospital, New York, New York

A 48,000-square-foot (4459-square-metre) site, containing liqueiable
soils within the depth range of 10–50 feet (3–15.2 metres) below grade,
was selected for the construction of a six-story structure at Harlem
322 Ground improvement

Figure 8.17 Compaction grout DAQ system with grout line under table where grout
pressure and volume is measured. (Courtesy of Hayward Baker Inc.)

Hospital in New York City. A combination of spread footings and a large

structural mat foundation on compaction grout–improved ground was
selected during a peer review as a value-engineered alternative to deep
foundations.
The grout injection pipe was installed and extracted in a continuous
operation utilising a large-tracked drill-mounted vibratory hammer (Figure
8.18). The geotechnical contractor designed and installed real-time data
acquisition systems on two rigs to record depth, grout pressures, grout vol-
umes, and grout injection rates. The plots of the data were produced in real
time to assist in QC. The operators had similar screen displays to allow
them to determine if criteria were being met and if they could advance to
the next stage. In addition, a three-dimensional visualization package was
created to assist in evaluation of the grouting process (Figure 8.19). The
three-dimensional rendering of grout volume and pressure was completed
automatically, based on data acquired from the ield.
Over 130 CPTs were performed to verify post-treatment soil improve-
ment. As each was performed, the electronic data was imported into a
programme which calculated the factor of safety against liquefaction, the
static settlement, and the seismically induced settlement (Figure 8.20).

8.10.2 Interstate 5 at Hasley Canyon, California

Compaction grouting was selected to densify liqueiable sands and gravels
around existing and newly installed piles supporting Interstate 5 where it
 Compaction grouting 323

Figure 8.18 Large track-mounted drill rigs with telescopic masts installed the injection
pipes in a single stroke to the maximum required treatment depth of 50 feet
(15.2 meters). (Courtesy of Hayward Baker Inc.)

crosses the Castaic Creek approximately 30 miles (48 kilometres) north

of Los Angeles, since it is capable of being performed from low headroom
working conditions. The creek is an ephemeral creek that is normally
dry. The soils in the creek bed are typical luvial sediments composed of
mostly cobbles, gravels and sand. SPT blow counts ranged from 8–15.
These soils would be prone to liquefaction during an earthquake. The
liqueiable soils were as deep as 30 feet (9.1 m) below ground surface.
The geotechnical contractor performed multiple stages of compaction
grouting to accommodate trafic requirements, sensitive habitat, limited/
restricted access, and construction sequencing along a major interstate
highway. There were approximately 100 injection locations to depths up to
59 feet (18 metres) grouted for abutment 1, 225 locations to depths ranging
from 52.5–59 feet (16–18 metres) grouted for pier 2, and more than 300
locations to depths of 16.4 feet (5 metres) for abutment 4.
A small crawler drill was used in the creek bed to drill grout pipes to
the depths required (Figure 8.21). Holes were drilled to depths ranging
from 15–59 feet (4.6–18 metres) below ground surface. Primary holes
were drilled on approximately 20-foot (6.1-metre) spacing, and secondary
holes were drilled between these holes leaving 10-foot (3-metre) spacing
 324
Ground improvement
0 300 600 900 1200 0 5 10 15 20 25 30 0 5 10 15 20
0 0 0

10 10 10

20 20 20

30 30 30

40 40 40

50 50 50

60 60 60

70 70 70
Max. grout pressure Grout volume(ft3) Grout rate (ft3/min.)
(psi) (1 ft3 = 0.03 m3)
(1 psi = 6.89 kPa)

Figure 8.19 Real-time plots of grouting data and three-dimensional visualization graphic. (Courtesy of Hayward Baker Inc.)
 qt [tsf] Rf [%] CSR and CRR Factor of safety Settlement [in]
0 200 0 2 0 2 0 2 0 2
0 0 0 0 0

10 10 10 10 10

20 20 20 20 20
Depth [ft]

Depth [ft]

Depth [ft]

Depth [ft]

Depth [ft]
30 30 30 30 30

40 40 40 40 40

Compaction grouting
50 50 50 50 50
CPT-4 CPT-4
CRR
CSR
60 60 60 60 60
Bottom of footing elevation = 7.5'' Work area 3-CPT 493N 1 tsf = 95.76 kPa Post-treatment seis. settlement = 0.0''
Approx. footing depth = 10' Harlem hospital, NY 1 ft = 0.3 m Pre-treatment seis. settlement = 3.2''
(16.5' × 16.5' footing) 10/13/08 1 in = 2.54 cm

325
Figure 8.20 Sample of pre- and post-treatment CPT and analysis results. (Courtesy of Hayward Baker Inc.)
326 Ground improvement

Figure 8.21 Small-tracked drill installing the injection pipes (left) and truck mounted
batch plant producing LMG (right). (Courtesy of Hayward Baker Inc.)

between holes. Five rows of holes were drilled along the pile lines. Figure
8.22 shows the grout layout and testing pattern for one of the bridge
abutments.
Once the casings were installed, compaction grout was injected in up-
stage fashion to densify the soil. Compaction grouting was conducted until
one of four criteria were met: (1) a given overpressure was achieved; (2) a
given quantity of grout was injected into the zone; (3) grout exited the
ground surface; and (4) ground or structural heave was observed. Once
refusal occurred, the casing was pulled up 2 feet (0.6 metres) and compac-
tion grouting continued. This process continued until the entire hole was
grouted.
SPT and CPT testing were conducted to verify that the soil had been suf-
iciently densiied following compaction grouting (Figure 8.23). The loca-
tion of the SPT or CPT test was determined by the owner, and was usually
midway between injection points. The acceptance criteria for the project
were to achieve minimum SPT blow count N160 of 36 or a CPT tip resis-
tance > 19 MPa.
A total of 766 holes were drilled, for a total length of 35,200 feet (10,732
metres). Grout consumption for the project was 53,208 cubic feet (1,507
cubic metres), or an average of 1.5 cubic feet (42 litres) per foot (0.3 metres)
of hole.

8.10.3 Tunnelling beneath Highway 85,

Denver, Colorado
A sinkhole opened up during rush hour on Highway 85 north of Denver due
to sloughing during a tunnelling operation. Within the hour, compaction
grouting crews were on site. Within hours the situation had been resolved
using grouting techniques.
 ABUT 4 PLAN VIEW
Test section 2
A Phase 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
Battered piles Battered piles
A
B TP-4
C C-32
D TP-6
TP-5 C-34 C-35 C-36 C-37 C-38 C-39 C-40 C-41 C-42
E
F 3m 3m
G C-33
3m TYP TYP
H TYP
J
of K 3m
TYP STA
Abut 4 L STA 902 + 26.207
M
2.31m
N C-51
P
Q
R C-52
S
T
U
V

‘A’ line

Compaction grouting
Critical phase 1 Critical Critical phase 1
phase 2
0.50 m
A’
Notes:
1. Primary grout hole location, inclined primary grout hole.
2. Secondary grout hole location, inclined secondary grout hole.
3. Pre-CPT location
4. Post-CPT location

Figure 8.22 Grout injection and testing layout pattern. (Courtesy of Hayward Baker Inc.)

327
328 Ground improvement

Figure 8.23 Compaction grouting densiies soils loosened during tunneling operations.
(Courtesy of Hayward Baker Inc.)

REFERENCES

American Society of Civil Engineers Standards 53-10-Consensus Guide Committee.

(2010). Compaction grouting consensus guide.
Byle, M.J. (1997). Limited mobility displacement grouting: When ‘compac-
tion grout’ is not compaction grout. In: Vipulanandan, C. (ed.), Grouting:
Compaction, Remediation and Testing, Geotechnical Special Publication No.
66, New York, New York: American Society of Civil Engineers, pp. 32–42.
Crockford R. and Bell A.L. (1996). Compaction grouting in the UK – a review. In:
Yonekura R., Terashi M., and Shibazaki M. (eds.), Grouting and Deep Mixing,
Proceedings of the Second International Conference on Ground Improvement
Geosystems, May 14-17, 1996, Tokyo, Japan: Balkema Publishers, pp.
279–284.
Graf, E.D. (1969). Compaction grouting technique, Journal of the Soil Mechanics
and Foundations Division, ASCE, Vol. 95, No. SM5, Proceedings Paper 6766,
pp. 1151–1158.
Henry, J.P. (1986). Low slump compaction grouting for correction of central Florida
sinkholes, National Water Well Association, Proceedings of the Conference
on Environmental Problems in Karst Terranes and Their Solutions, October
28–30, 1986, Bowling Green, Kentucky, United States.
Hussin, J.D. and Ali, S. (1987). Soil improvement at the Trident submarine facility.
In: Welsh, J.P. (ed.), Soil Improvement – A 10-Year Update, Proceedings of
a Symposium, ASCE Convention, April 28, 1987, Atlantic City, New Jersey,
United States, pp. 215–231.
Warner, J. (1982). Compaction grouting – the irst thirty years, Proceedings of the
ASCE Specialty Conference, Grouting in Geotechnical Engineering, February
25–28, 1982, New Orleans, Louisiana, United States, pp. 694–707.
Chapter 9

In-situ soil mixing

Michał Topolnicki

CONTENTS

9.1 Introduction ................................................................................. 330

9.2 Historical development and classiication ..................................... 332
9.3 Equipment and execution ............................................................. 342
9.3.1 Dry method deep mixing ................................................... 342
9.3.2 Dry method shallow mixing .............................................. 345
9.3.3 Wet method mechanical deep mixing ................................ 351
9.3.3.1 Mixing tools rotating about vertical axes............. 351
9.3.3.2 Mixing tools rotating about horizontal axes ........ 358
9.3.3.3 Vertical trenching ................................................. 360
9.3.4 Wet method mechanical shallow mixing ........................... 362
9.3.5 Wet method hybrid deep mixing........................................ 364
9.3.6 Installation process (mixing about vertical axis)................ 367
9.3.7 Details of construction and execution (mixing about
vertical axis) ...................................................................... 370
9.3.7.1 Number of shafts.................................................. 370
9.3.7.2 Shape and orientation of mixing blades................ 371
9.3.7.3 Position of injection nozzles.................................. 372
9.3.7.4 Degree of mixing.................................................. 372
9.3.7.5 Control of binder supply....................................... 374
9.3.7.6 Control during construction................................. 375
9.4 Applications and limitations......................................................... 376
9.4.1 Areas of application ........................................................... 376
9.4.2 Patterns of deep soil mixing installations........................... 376
9.4.3 Foundation support ........................................................... 379
9.4.4 Retention systems .............................................................. 381
9.4.5 Ground treatment .............................................................. 383
9.4.6 Liquefaction mitigation...................................................... 384
9.4.7 Hydraulic cut-off walls ...................................................... 385
9.4.8 Environmental remediation ............................................... 386
9.4.9 Advantages and limitations................................................ 388

329
330 Ground improvement

9.5 Design .......................................................................................... 388

9.5.1 General design procedure .................................................. 388
9.5.2 The choice between dry and wet process of soil mixing .... 390
9.5.3 Engineering properties of stabilised soil............................. 392
9.5.4 Long-term strength gain and deterioration of
stabilised soil ..................................................................... 398
9.5.5 Design strength of stabilised soil........................................ 399
9.5.6 Geotechnical design........................................................... 403
9.6 Quality assessment of deep mixing............................................... 407
9.7 Selected case histories ................................................................... 413
9.7.1 Metropolitan intercity expressway (dry DM)..................... 413
9.7.2 Road embankment (dry mass stabilisation and dry DM) ... 414
9.7.3 Carriageway Trasa Zielona (wet DM)................................ 416
9.7.4 Deepwater bulkhead in soft soils (wet DM) ....................... 417
9.7.5 Foundation of a multistory building (wet DM) .................. 418
9.7.6 Foundation of highway bridge supports (wet DM) ............ 420
9.7.7 Excavation protection (wet DM, reinforced columns)........ 422
9.7.8 Cut-off block to install a crosswise connection between
two tunnels (CSM, wet DM) ............................................. 423
9.7.9 Excavation control (hybrid wet mixing)............................. 425
References ............................................................................................. 426

9.1 INTRODUCTION

The use of soil mixing (SM) to improve the engineering and environ-
mental properties of soft or contaminated ground has increased widely
since its genesis. Growing interest for SM mainly results from the high
lexibility of this method, which can be purposely adapted to speciic
project requirements and site conditions, as well as from cost-to-per-
formance eficiency of respective geotechnical solutions. In this method
of ground improvement, soils are mixed in situ with different stabilis-
ing binders, which chemically react with the soil and/or the groundwa-
ter. The stabilised soil material that is produced generally has a higher
strength, lower permeability, and lower compressibility than the native
soil. The improvement becomes possible by cation exchange at the sur-
face of clay minerals, bonding of soil particles, and/or illing of voids
by chemical reaction products. The most important binders are cements
and limes. However, blast-furnace slag, gypsum, and ashes as well as
other secondary products and compound materials are also used. For
environmental treatment, binders are replaced with chemical oxidation
agents or other reactive materials to render pollutants harmless.
 In-situ soil mixing 331

Soil mixing can be subdivided into two general methods: the deep mix-
ing method (DMM) and the shallow mixing method (SMM). Both DMM
and SMM include a variety of proprietary systems.
The more frequently used DMM is applied for in-situ stabilisation of the
soil to a minimum depth of 3 m (a limit depth introduced by EN 14679:2005).
The binders are injected into the soil in dry or slurry form through hollow
mixing shafts tipped with various cutting tools and equipped with auger
lights, mixing blades, or paddles to increase the eficiency of the mixing
process. The shafts, mounted in single or multiple arrangements, rotate
about the vertical axis and produce individual or overlapping soil-mix
columns. In the case of special cutting/mixing arms equipped with cut-
ter wheels mounted on horizontal axes to create panels or with a revolv-
ing cutting chain to create continuous walls, the slurry is pumped through
injection pipes and outlets mounted along the arm. In some methods, the
mechanical mixing is enhanced by simultaneously injecting luid grout at
high velocity through nozzles in the mixing or cutting tools.
The complementary SMM has been specially developed to reduce the
costs of improving loose or soft supericial soils overlying substantial areas,
including land disposed dredged sediments and wet organic soils a few
metres thick. It is also a suitable method for in-situ remediation of contami-
nated soils and sludges. In such applications, the soils have to be thoroughly
mixed in-situ with an appropriate amount of wet or dry binders to ensure
stabilisation of the entire volume of treated soil. Therefore, this type of soil
mixing is often referred to as ‘mass stabilisation’. Mass stabilisation can
be achieved by installing vertical overlapping columns with up and down
movements of rotating mixing tools, as in the case of standard DMM, and
is most cost-effective when using large-diameter mixing augers or multiple
shaft arrangements. For shallow depth applications, however, generally
limited to about 5 m, another very eficient method of mass stabilisation is
usually implemented, and the mixing process is carried out repeatedly in
vertical and horizontal directions through the soil mass using various cut-
ting and mixing tools that are different from the tools developed for DMM
(e.g., mixing drums). Consequently, in the classiication scheme used in this
chapter the SMM includes both systems of mass stabilisation.
In-situ soil mixing is a versatile ground improvement method. It can be
used to stabilise a wide range of soils, including soft clays, silts and ine-
grained sands. Stabilisation of organic soils such as gyttja (sedimentary
organic soil), peat, and sludges is also possible, but is more dificult and
requires carefully tailored binders and execution procedures. However, the
engineering properties of the stabilised soil will not only depend on the
characteristics of the binder. They will also depend, to a large extent, on
the inherent characteristics of each soil and the way it has been depos-
ited, as well as on mixing and curing conditions at a particular worksite.
Therefore, a thorough understanding of chemical reactions with these
332 Ground improvement

factors is necessary in order to ensure successful application of this ground

improvement technology.
In this chapter, the current status of in-situ soil mixing is outlined, taking
into account recent execution and design practice, international literature,
and experience. General application areas are identiied and discussed, and
a few case histories selected from international projects are included for
illustration. The focus is on civil engineering applications of DMM, and,
to a lesser extent, of SMM. Some specialised soil mixing issues in relation
to environmental projects, such as mass treatment of subsurface hazardous
wastes by various processes including solidiication, stabilisation and chemi-
cal treatments, reactive barriers, etc., are only touched upon, therefore the
cited literature should be referred to for more information. Furthermore,
overly extensive descriptions of the complicated chemical processes occur-
ring in the stabilised soil when mixed with various binders have been
excluded from the contents. This choice, however, should not undermine the
importance of this aspect of soil mixing. It may rather relect the fact that
in spite of considerable knowledge about basic reaction mechanisms, identi-
ied and described for instance by Babasaki et al. (1996) for soils stabilised
with lime or cement, it is still not possible to predict the strength of in-situ
mixed soil with a reasonable level of accuracy. As a consequence of this
fundamental deiciency, which we are challenged to overcome, it is believed
that the development of SM will be continued along a somewhat erratic
experimental path, and will be to a large extent dependent on accumulated
experiences. Therefore, the scope of this chapter instead concentrates on the
characteristics of equipment in current use, execution procedures with refer-
ence to selected operational methods, applications, merits, and the limita-
tions of the technology. Design aspects as well as quality control and quality
assurance issues of DMM are also considered. The design approach out-
lined herein follows the practice established in Japan, the US, and Europe,
assuming that the treated soil is practically an impermeable material. The
approach used with respect to DM columns stabilised with unslaked lime or
lime and cement, which may act as vertical drains, is covered in Chapter 10.

9.2 HISTORICAL DEVELOPMENT

AND CLASSIFICATION

The historical development of SM was extensively covered in the second edi-

tion of this book, taking into account the status of this technology about 10
years ago. Anticipated and observed growth of applications worldwide has
signiicantly changed this situation, however, and SM can now be regarded
as well-established ground improvement technology. Consequently, rel-
evant information on historical development has been shortened to include
only the most important and stimulating achievements.
 In-situ soil mixing 333

The roots of deep soil mixing go back to the mid-1950s, when the
mixed-in-place (MIP) piling technique was developed by Intrusion-Prepakt
Inc. (FHWA, 2000). In this method a mechanical mixer was used to mix
cementitious grout into the soil for the purpose of creating foundation ele-
ments and retaining walls. The grout was injected from the tip of a mixing
tool consisting of a drilling head and separated horizontal blades. Modern
deep mixing techniques relect, however, mainly Japanese and Scandinavian
efforts over the last ive decades as well as recent European achievements in
cutter wheel and trench-type mixing systems.
The level of research and development activity in Japan in relation to
DMM remains the highest in the world. The irst commercial application
of the deep lime mixing (DLM) method, utilising a mechanical binder
feeding system, was conducted in 1974 by Fudo Construction Co. Ltd.
using the Mark IV machine to improve reclaimed soft alluvial clay in
Chiba Prefecture in Japan. The irst marine use of DLM was in 1975 at
Tokyo Port (Terashi, 2002a). In an effort to improve the uniformity of
the stabilised soil, a new concept using cement mortar and cement-water
slurry as binders was implemented in the mid-1970s, with CMC and DCM
(deep cement mixing) methods developed by Kawasaki Steel & Fudo and
Takenaka Group, respectively, with close supervision from Port Harbour
Research Institute. The irst on-land and marine applications of CMC and
DCM were conducted in 1976. Also that year, the Seiko Kogyo Co. devel-
oped and introduced the soil-mixed wall (SMW) method using discontinu-
ous augers and paddles positioned at discrete intervals, usually along three
shafts arranged in a row. This method was applied primarily for excavation
support and groundwater cut-off walls, with the possibility for installation
of reinforcing steel sections within fresh columns to increase bending stiff-
ness of the supporting DM elements.
Major marine ground improvement works at Daikoku Pier, beginning in
1977 and continuing for about 10 years, contributed to important develop-
ments of the wet method of deep mixing (e.g., DCM, DECOM, POCOM,
and others). These developments included the elaboration of design stan-
dards and construction control procedures, slowly hardening binders and
new positioning systems for offshore applications (Terashi, 2002a).
A general method using a variety of stabilising binders in slurry form
(wet method) has been named cement deep mixing (CDM) method. In
1977, the CDM Association was established in Japan to promote and
improve the CDM method via a collaboration of general contractors,
marine works, and foundation works contractors, as well as industrial
and research institutes. As a result, new eficient machines were developed,
such as CDM-Mega, CDM-LODIC, CDM-Land4, and CDM-Column
21. For marine applications the CDM method has mainly been used to
improve the foundations of revetments, as well as quay wall and breakwa-
ter foundations. The diameter of the mixing blades ranges from 1.0–1.6 m
334 Ground improvement

and the maximum depth of improvement is about 70 m below water. For

land applications, the CDM method has been mainly applied for slide and
liquefaction prevention, settlement reduction, and to improve the bearing
capacity of foundations. The standard CDM machines have two shafts,
mixing blades with a diameter of 1 m, and a penetration depth limited to
about 50 m. Typical machines for marine and on-land use are shown in
Figures 9.1 and 9.2, respectively.
Another remarkable development conceived in Japan in 1993 is the TRD
(trench re-mixing and cutting deep wall) method. In this system, a continu-
ous soil-mix wall is created in situ by lateral motion of vertical ‘chain saw’
in a one-phase process that involves simultaneous full-depth cutting and
mixing of soils with binders in slurry form. This method has been applied
in Japan for more than 400 projects, and in the US since 2006 (Figure 9.3).
Approximately two thirds have been structural retaining walls and one
third were cut-off walls (Garbin et al., 2010).
The development of the wet method in Japan includes successful attempts
to combine mechanical mixing with high-velocity injection. In 1984, the
spreadable wing (SWING) method was introduced. In this unique system
a retractable mixing blade mounted on a single drilling shaft allows treat-
ment of speciic depths with large diameters (0.6 m with blade retracted and
up to 2 m after expansion). Following that, jet grouting was incorporated
into SWING and its irst application was in 1986. With additional jetting
during withdrawal, mechanically mixed and jet mixed concentric zones
are produced with a total diameter up to 3.6 m (Kawasaki et al., 1996).

(a) (b) 830

830
830

∅1000
830 mm

Figure 9.1 (a) CDM barges for marine deep mixing works using the wet method, Japan.
(Courtesy of CDM Association, Japan.) (b) Possible arrangement of eight
mixing shafts.
 In-situ soil mixing 335

(a) (b) (c)

Figure 9.2 CDM machine for on-land deep mixing works using the wet method, Japan
(note auger screws along the shafts enabling partial extrusion of disaggre-
gated soil).

(a) (b)

Figure 9.3 (a) Execution of a cut-off wall with the TRD method (Herbert Hoover Dike
rehabilitation project in Florida), (b) close-up of the cutting/mixing chain.
(Courtesy of Hayward Baker, Inc.)
336 Ground improvement

Moreover, in 1992, the jet churning system management (JACSMAN)

method using combined mechanical mixing and modern cross-jet grouting
systems was developed by Fudo Construction Inc. and Chemical Grouting
Co. (Kawanabe and Nozu, 2002). Another innovative development of Fudo
is the CI-CMC method. With this method it is possible to disperse solidify-
ing materials to the entire improvement area by injecting cement slurry into
the air stream with an ‘ejector discharge’ device incorporated into the stir-
ring blade. Consequently, penetration loads are reduced and the soil becomes
easier to mix due to the air lifting effect, resulting in better homogeneity,
reduced lateral displacements and lower costs of mixing (Fudo Inc., 2011).
In 1978 the development study of Japanese dry method designated for
on-land applications utilising a pneumatic binder feeding system known
as dry jet mixing (DJM) was initiated. The constructed machine was irst
applied in 1981. It was subsequently improved, with a landmark project
in 1985 on San-yo Motorway, where a 10-m-high embankment was con-
structed on 10 m of sensitive soft clay (Terashi, 2002a). In the early 1980s
the DJM Association was established in Japan, with a role similar to that
of the CDM Association for the wet method. The current standard DJM
equipment has two mixing shafts, with blades of 1.0 m diameter and a
maximum penetration depth of 33 m, as shown in Figure 9.4.

(a) (b)

Figure 9.4 (a) DJM machine for on-land deep mixing works using the dry method, Japan.
(b) Two mixing tools, diameter 1.0 m.
 In-situ soil mixing 337

The development of DMM in Scandinavia was initiated in Sweden in

1967 when laboratory and ield research began for a new method of sta-
bilisation of soft clays with unslaked lime. The irst light wheel-mounted
mixing equipment was manufactured by Linden-Alimak AB in coopera-
tion with Swedish Geotechnical Institute, BPA Byggproduktion (presently
LCM AB, Keller Group), and Euroc AB. In Finland, research was initi-
ated during that time as well. Commercial use of the lime column method
started in Sweden in 1975 for support of excavation, embankment stabili-
sation, and shallow foundations near Stockholm. Other types of dry bind-
ers, like cement and two component binders composed of unslaked lime
and cement, have been subsequently investigated and put into practice. The
irst commercial project with the lime cement column method in Finland
was conducted in 1988 and then 1990 both in Sweden and Norway under
Swedish guidance and using Swedish contractors. This type of DMM,
generally with two component binders tailored for various soft soils, has
been consequently developed over the years. It is now referred to as the
‘Nordic dry deep mixing method’ (Holm, 2002). With an increasing num-
ber of proven applications, especially since 1989, by the mid-1990s this
method had become the predominant technology of ground improvement
in Scandinavia. Relatively light and mobile equipment with one shaft is
typically used to produce columns of 0.5–1.0 m in diameter to a maximum
depth of about 25 m depending on the soil conditions. The application
focus remains on ground improvement to reduce settlement and enhance
stability of road and railroad embankments, and soil/column interaction
solutions for very soft, highly compressible clayey and/or organic soils. The
irst application for mitigation of vibrations induced by high speed trains
took place in Sweden in 2000 (Holm, 2002a).
In European countries outside Scandinavia, the Nordic method has been
mostly used in Poland since 1995, with a recent major application during
construction of road S7 near Elbla˛ g, involving about 743,000 lin. m of dry
DM columns (Figure 9.5). Increasing number of projects were also conducted
in the UK (irst in 2001), and ield trials in The Netherlands and Germany.
In Finland another dry shallow mixing method for stabilisation of super-
icial layers of peat, mud, sludge, or soft clay to a depth of about 5 m has
been developed since 1992 and applied for road and land reclamation proj-
ects. The mixing tools of this method of mass stabilisation have different
shapes and are typically attached to the arm of a conventional excavator.
They may be constructed as mixing/cutting heads, equipped with blades
rotating about a vertical or a horizontal axis, or as mixing drums. The
mixing process is conducted repeatedly in vertical and horizontal direc-
tions through the soil mass in order to obtain a homogeneous soil-binder
mixture. The irst commercial project was conducted in Sweden in 1995 in
connection with renovation works along Highway 601 Sundsvägen, where
about 10,000 m3 of peat were treated.
338 Ground improvement

(a) (b)

Figure 9.5 Dry method deep mixing near Elbla˛g (Poland) using the Nordic method.
(a) Keller/LCM equipment. (b) Dry mixing in winter.

The application and development of the contemporary DMM in the

United States started in the mid-1980s and was comprised initially of
the wet method. In 1986, SMW Seiko Inc. began operations under license
from Japanese parent Seiko Kogyo Co. The SMW method was subsequently
used in 1987–1989 in a landmark liquefaction mitigation and seepage cut-
off project at Jackson Lake Dam, Wyoming, US, where 130,000 lin. m of
column were installed to a maximum depth of 33 m (FHWA, 2000).
Following their cooperation with SMW Seiko on the Jackson Lake
Dam project, in the late 1980s Geo-Con Inc. developed the irst US
soil mixing technologies, the DSM and SSM (shallow soil mixing) meth-
ods. The DSM method uses 1–6 shafts with discontinuous augers of
0.8–1 m in diameter. The SSM method uses a large-diameter single mixer
to economically treat weak supericial soils and contaminated sites to a
depth of about 12 m. The SSM has since been extended to accommodate
binders in a dry form. This variant using dry form binders was applied
in 1991 to stabilise large lagoons containing contaminated sludge resi-
dues from a water treatment plant at a reinery near Chicago (Jasperse and
Ryan, 1992).
The Japanese SCC method was introduced in the US by SCC Technology
Inc. The single-axis system of Hayward Baker (Keller), with diameters of
0.5–2.5 m, typically 2.1 and 2.4 m, began development in the 1990s and
has been applied since 1997 (Burke, 2002). Mass stabilisation with the wet
method (shallow soil-cement mixing) was also applied, using an excavator
equipped with a shallow mixing bucket (Druss, 2002). The bucket con-
tained mixing blades that rotate about a horizontal axis.
The largest DM works in North America to date were conducted between
2009 and 2011 for the New Orleans East Back Levee (LPV-111). The proj-
ect consisted of ca. 1.3 million cubic metres of wet DM to a depth of up
to 20.5 m. Over 17,000 single and double axis elements of diameter 1.6 m
 In-situ soil mixing 339

were combined in different patterns to create transversal panels, approxi-

mately 4.7 m apart, along the 8.53 km extension of the levee. Eight batch-
ing plants were used to prepare the grout using over 460,000 tonnes of slag/
cement binder (Schmutzler et al., 2011).
A combined mechanical and hydraulic mixing method called GEOJET
(Condon Johnson and Associates, Halliburton) has been developed and
modiied in the US since the early 1990s. GEOJET equipment includes the
soil processor equipped with specially designed cutting blades and multiple
jetting nozzles which jet mix at pressures up to 35 MPa. The irst commer-
cial application was in 1994, followed by some major retaining wall works
and installations of pipe piles in soil.
The irst commercial project that used the Nordic method in the US
was conducted in 1996 in Queens, NY, by the Stabilator Company
(Skanska). Subsequent application for settlement reduction at I-15 in Salt
Lake City, Utah, took place in 1997. Since 1998, other dry methods, like
DJM (Raito Inc.) and TREVIMIX (Treviicos Corp., with Hercules), have
been available.
In Europe, the earliest wet DM activities that took place in the 1980s were
oriented towards development of a potentially cheaper alternative to jet
grouting. In France, Bachy Soletanche developed the COLMIX method in
the mid-1980s, in conjunction with the French Railway Authority (SNCF)
and the French National Laboratory for Roads and Bridges (LCPC). The
method features twin, triple, or quadruple contra-rotating and interlocking
augers, generally 3–4 m long and driven via hollow stem rods coupled to a
single rotary drive. Several road and rail embankment stabilisation projects
have been completed with this method in France, UK, and Italy, as sum-
marised by Lebon (2002).
In Germany, the irst application of the mixed-in-place (MIP) system
developed by Bauer Spezialtiefbau GmbH, which was based on the rotary-
auger-soil-mixing (RASM) method utilising single shafted crane and wet
binder, took place in Nürnberg in 1987. MIP piles were executed to cre-
ate panels of mixed soil illing up a ‘Berlin’-type temporary retaining
wall constructed in sands (Herrmann et al., 1992). Subsequently, a more
advanced triple auger wet mixing system has been developed since the early
1990s. This system has been in use since 1994, primarily for construction
of temporary and permanent panels supporting excavations, cut-off walls,
ground improvement, and environmental purposes (Außenlechner et al.,
2003, Schwarz and Seidel, 2003). For walls with shallow depth, typically
6–15 m, the Bauer soil mixing wall method, which uses three adjacent
slightly overlapping augers and mixing paddles, was later developed. Keller
Grundbau GmbH developed their irst system based on a single paddle
shaft equipped with a short auger and mixing blades above the drill bit.
Their commercial ground improvement applications for this system have
been ongoing since 1995. More advanced mixing tools with twin and triple
340 Ground improvement

shafts as well as combined systems involving mechanical and jet assisted

mixing have been also introduced.
Another high-capacity specialised wet mixing system developed in
Germany in 1994 is the FMI method (Fräs-Misch-Injektionsverfahren =
cut-mix-injection). It was applied for the irst time in 1996 in Giessen
(Pampel and Polloczek, 1999). The FMI machine is comprised of a spe-
cial cutting tree, along which cutting blades are rotated by two chain sys-
tems. The cutting tree can be inclined up to 80 degrees, and is dragged
through the soil behind the power unit. With this method it is possible
to treat the soil in deep strips, with a mean capacity of 70–100 m3/h.
The width of treatment is 1 m down to a depth of 6 m, or 0.5 m down
to a depth of 9 m. Initial applications mainly covered ground improve-
ment works along railways. A similar system, called TRENCHMIX, was
jointly developed by Bachy Soletanche and Mastenbroek and irst applied
in France in 2005.
In the United Kingdom wet DM for ground improvement was employed
in early 1990s by Cementation Piling and Foundations for construction
of a few temporary shafts of approximately 4 m internal diameter and
up to 15 m deep (Blackwell, 1994). The columns were installed with a
simple auger-type mixing tool, using ive passes of the tool over a 1 m
withdrawal length. Around 1995, soil mixing was introduced for geoen-
vironmental applications, with growing importance since 1997. Currently
the UK is leading Europe in the research and application of wet mixing to
the containment and encapsulation of contaminated soils, including cut-
off walls and reactive barriers (Lebon, 2002, Al-Tabbaa et al., 2009). In
2001, the UK saw its irst use of the dry Nordic method (Keller), and in
2010 the TRENCHMIX machine was irst used for lood defence works in
Nottingham (Bachy Soletanche).
In Italy, the Trevi SpA developed in the late 1980s a dry mixing method
named TREVIMIX. The equipment has more similarities with the Japanese
DJM method than with the Nordic method. In this system one or two
(more common) shafts with mixing paddles of 1.0 m (or 0.8 m) in diameter
are arranged at variable spacings of 1.5–3.5 m and are used to disintegrate
soil structure during penetration with air. The distinction of this system lies
in its ability to operate in dry or semi-dry conditions by adding a controlled
amount of water to the soil in order to ensure a hydrating reaction. First
applications in Italy have been reported by Pavianni and Pagotto (1991).
Another development is the TURBOJET wet mixing system that uses a
tubular Kelly with drilling bit and mixing blades, and combines mechani-
cal mixing and single luid jet grouting technology.
In Poland, the wet method of DM was irst introduced in 1999 by Keller
Polska, initially using single-axis equipment and later twin-shaft tools. The
irst project involved execution of intersecting columns forming a cut-off
wall along an old dam of the Vistula River in Kraków. Since then, the use
 In-situ soil mixing 341

of wet DMM in Poland is probably the highest in Europe (except for envi-
ronmental applications), including the irst worldwide applications of DM
for the foundations of highway bridges (irst in 2002) and modern wind
turbines (irst in 2007).
In Belgium Smet-Boring NV has developed a modiied DM system, called
TSM (Tubular Soil Mixing). TSM uses a mixing auger inside an outer cas-
ing, diameter 43–63 cm, equipped with a set of nozzles for high-pressure
jetting. The casing contains holes at its bottom section in order to avoid
blockage in the ground if pressure in the soil-slurry mixture builds up.
The casing also reduces the lateral displacement of soil/slurry and improves
homogeneity of soil-mix. The major applications are for excavation sup-
port, with a rather small column overcut (about 5 cm) because of a high
accuracy of vertical alignment.
Another important European achievement is the cutter soil mixing (CSM)
system, derived from the cutter diaphragm walling technique, jointly devel-
oped by Bauer Maschinen GmbH and Bachy Soletanche since 2003. The
soil is broken down and mixed in situ with slurry by two sets of cutting/
mixing wheels rotating about horizontal axes (cf. Fiorotto et al., 2005).
The CSM system offers signiicant advantages over other walling tech-
niques, and has been successfully transferred to countries outside Europe,
including Japan and the US.
Deep SM is also very popular in China and Southeast Asia. In China
DMM and DJM were introduced in 1977 and 1983, respectively (Zheng
and Liu, 2009). Both methods are now widely used for a variety of applica-
tions and proved to be competitive in terms of costs and time of execution.
To improve the performance of DM columns various types of composite
columns have been also developed (CDMC). They may include installation
of a precast concrete pile, reinforced concrete pile, or a steel pipe inside a
fresh DM column, or even execution of DM column inside a sand column
of comparatively large diameter. In Korea DCM has been developed since
mid-1980s and the use is increasing, especially for marine and harbour
works using special barges equipped with multiple mixing shafts (Kim
et al., 2009).
The hitherto development of different technologies and equipment
used in SM is dificult to follow without a certain generic classiication
system. Several similar systems have already been developed for this pur-
pose (FHWA 2000, CDIT 2002, and EN 14679). The classiication format
adopted herein is based on three fundamental operational characteristics.
The distinction between wet and dry technologies with respect to the form
of binder introduced into the soil is the most straightforward, and hence
the most widely used format. In the dry mixing methods the medium for
binder transportation is typically compressed air, while in the wet mix-
ing methods the medium of transportation is typically water. The second
key characteristic is related to the method used to mix the binder (i.e., by
342 Ground improvement

mechanical action of the mixing tool with the binder injected at relatively
low velocity), hydraulic action of the luid grout injected at high velocity
(jet grouting), or by a combination of both aforementioned techniques (so-
called hybrid mixing). The third basic characteristic relects the location
of mixing action at the end or along a speciic tool. The elaborated clas-
siication chart with the allocation of several operational methods, split
with account for the difference between systems involving rotation of the
cutting/mixing tool about vertical or horizontal axes or around the whole
cutting arm, is shown in Figure 9.6.
When comparing technical features of recently used DMM and SMM
machines and operational systems, it should be kept in mind that the afore-
mentioned methods have been developed while taking into account various
demands and constraints of regional markets, as well as soil conditions
prevailing in areas of potential application. Moreover, various operational
systems also relect different objectives of ground improvement and design
approaches. Consequently, not all SM methods can be regarded as equiva-
lents, although all are based on the same overall concept of in-situ soil
stabilisation. Despite these variations, the main technical goal of any SM
method is to ensure a uniform distribution of binder throughout the treated
soil volume, with uniform moisture content, and without signiicant pock-
ets of native soil or binder.

9.3 EQUIPMENT AND EXECUTION

9.3.1 Dry method deep mixing

Typical dry method DM construction equipment consists of a station-
ary or movable binder storage/premixing and supply unit, and a mixing
machine for the injection of binder material and installation of the col-
umns. The binder is delivered to the mixing machine by compressed air.
The equipment components generally include: silos with stabilising agents,
pressurised tank with binder feeder system, high-capacity air compres-
sor, air dryer, ilter unit, generator, control unit, and connecting hoses.
The two major techniques for dry mixing are the Japanese DJM and the
Nordic method.
The DJM mixing machines are equipped with one or two mixing shafts,
and are able to install columns to a maximum depth of 16–33 m (DJM
Assoc., 2002). A dual mixing shaft is the current standard outit, while a
single shaft may be used in narrow working areas or for sites with head-
room restrictions. The driving unit of the mixing shafts is located at the
foot of the tower to improve machine stability while the shafts are kept
together with a transverse steel bar, allowing for interlocking or tangential
positioning of the mixing blades. The bar, and sometimes additional freely
 In-situ soil mixing 343

In-situ soil mixing methods

(to be identified along arrows only)

Dry mixing Wet mixing

a (dry binders) a (binders in slurry form)

Mechanical Mechanical + Jet

b mixing b Jet mixing b mixing

At the end of mixing tool Along tool At the end of mixing tool
c c c

Vert. axis rot. Vert. axis rot. Vert. axis rot. Vert. axis rot. Vert. axis rot.
- DJM Assoc. - CDM Assoc. - SMW - SWING (out of scope of
(Japan) (Japan) (Japan, USA) (Japan) this chapter)
- Nordic method - SCC - DSM - JACSMAN, Systems:
(Sweden, (Japan) (USA) CI-CMC - Single
Finland) - SSM - MULTIMIX (Japan) (grout)
- TREVIMIX (USA) (Italy, USA) - GEOJET - Double
(Italy) - Keller system - COLMIX (USA) (air+grout)
- SMM: Mass (USA, Europe) (France) - HYDRAMECH - Triple
stabilisation - MECTOOL - Bauer Triple (USA) (water+air
(Finland, Swe- (USA) Auger System - TURBOJET +grout)
den, USA) ..... (Germany) (Italy) - Super jet
.... ..... - Cross jet
....
Horiz. axis rot. Horiz. axis rot. Vert. trenching Horiz. axis rot.
- SMM: Mass - CSM - FMI - CT Jet
stabilisation (Cutter soil (Germany) (Cutter turbo
with mixing mixing, - TRD jet, Italy)
drums Germany, (Japan)
(Finland) France) - TRENCHMIX
(France)

Figure 9.6 General classiication of in-situ soil mixing based on (a) binder form, (b) mixing
principle, and (c) location of mixing action, with allocation of selected fully
operational methods developed in various countries, split with respect to
rotation characteristic of the cutting/mixing tool.
344 Ground improvement

rotating (undriven or counteracting) mixing blades, also function to pre-

vent rotation of soil adhering to the driven mixing blades and shaft. The
standard mixing tool has a diameter of 1 m and consists of two full-length
mixing blades, mounted at the end of the shaft at two different levels, with
90° shift (Figure 9.7). A recently implemented modiied version of the tool
has a 1.3-m diameter (Aoi, 2002). To prevent choking, the injection ports
are positioned at the mixing shaft, below and behind angled mixing blades.
The lower port is used to inject air during penetration, or air and binder in
the case of soils requiring a high amount of stabilising agent. The binder is
mostly injected through the upper port into a cavity space created by the
mixing blade during withdrawal of the shaft with reversed rotation. Binder
quantity is adjusted by changing the rotation speed of the feeding wheel.
Air pressure and the amount of binder are automatically controlled to sup-
ply the speciied dosage of binder to the treated zone of soil. A hood cover-
ing the mixing tools is lowered to the soil surface to suppress dust emission
during work, while a square mixing shaft is generally used to facilitate
easier expulsion of injected air from the ground.
With torque capacity in the range of 20–30 kNm, the DJM machines are
able to conduct mixing operation in stratiied soils with varying resistance
(rotation is agitated by hydraulic or electric motors). The limits for execu-
tion are 70 kPa maximum shear strength for stiff clays and SPT N-value of
15 in sands (Terashi, 2003). Typical penetration speed in soft soil is 1–1.5
m/min with 24 or 32 rpm (electric motors) and an air low rate of 2 m3/min
to prevent choking of injection ports. During withdrawal (with counter-
rotation) the speed is typically 0.7 or 0.9 m/min, with 48 rpm or 64 rpm,

(a) (b)

Upper mixing
blade

Section of
Nozzle
the shaft

Lower mixing Air gathering

blade fins

Cutting bit

Figure 9.7 Mixing tools of the DJM method. (a) Construction scheme. (Redrawn from
DJM Association, 2002.) (b) Recently used single mixing tool of 1.0 m diameter.
 In-situ soil mixing 345

respectively (electric motors), and air low rate of 5 to 3 m3/min at shal-

low depths. Consumed air volume may vary between 2–9 m3/min, requir-
ing heavy-duty compressors with a capacity of 10.5–17 m3/min/shaft. The
volume of the binder tank is usually 2–3.5 m3 per one mixing shaft (DJM
Assoc., 2002).
The mixing machines developed in Sweden and Finland are lighter than
the Japanese rigs and are equipped with one mixing shaft. They are con-
structed to work mainly in soft to very soft soils with undrained shear
strength below approximately 25 kPa (maximum 50 kPa). The torque
capacity at 180 rpm is typically about 7 kNm, and increases to 30–40 kNm
at 20–30 rpm (some machines have two engines driving the Kelly rotation).
This allows for the installation of 0.6–0.8 m diameter columns to a depth
of about 25 m, 1–5 m from the edge of the base unit. The columns can be
also inclined up to about 1:4; maximum 1:1.
The equipment on site usually consists of a drill rig and a separate self-
driven mobile shuttle, hosting pressurised binder material tank(s), an air
dryer and a compressor. In addition to the plant items working on the con-
struction site, there is usually a requirement for a premix station including
a ilter unit, especially when delivery of ready-to-use binder is too expen-
sive. The binder shuttle moves between the premix station and the drill rig,
which is normally working several hundred metres away from the premix
station. During production, the shuttle is connected to the drill rig by an
umbilical through which the binder passes (via compressed air), along with
monitoring information on the binder mixing and supply rate. For shal-
low penetration depths, combined mixing machines with on-board instal-
lations are available. The amount of discharged binder is controlled with a
cell feeder mechanism, located at the bottom of the supply tank.
The air containing the binder is transported through the hollow Kelly bar
to an exchangeable mixing tool, mounted at the end. Typical mixing tools
consist of horizontal and curved or angled cutting/mixing blades, as shown
in Figure 9.8. The injection outlet is located at the central shaft, close to
the upper horizontal mixing blade. After the required depth is reached, the
mixing tool is lifted and simultaneously rotated in reverse, while the binder
material is horizontally injected to the soil. Typical withdrawal speed is
15–25 mm per rotation, with about 150–180 rotations per minute.
A summary of mixing conditions for selected dry DM methods is pre-
sented in Table 9.1.

9.3.2 Dry method shallow mixing

Shallow dry method mixing offers a cost-effective solution for ground
improvement works or site remediation when dealing with substantial vol-
umes of very weak or contaminated supericial soils with high water con-
tent, such as deposits of dredged sediments, wet organic soils, or waste
346 Ground improvement

(a) (b)

(c) (d)

Figure 9.8 Selected mixing tools. (a) SD 600 mm, (b) modiied SD 600 mm, (c) PB3 600
mm, (d) peat bore 800 mm. Note: changed location of binder outlow hole in
relation to the horizontal mixing blade in standard (a) and modiied (b) tool.
(Courtesy of LCM.)

sludges. In-situ mixing of the encountered soil mass with dry reagents to
the depth of a few metres can be economically carried out with large-diam-
eter single-axis augers, or with recently developed mass mixing tools imple-
mented in Finland and Sweden. In such applications it is also quite common
for the topsoil to be too weak to provide safe support for heavy mixing
machines. Therefore, it is best to use execution methods that employ mix-
ing tools suspended from the crane or mounted on elongated cantilever
arms, as they usually offer more lexible operation in the ield.
The shallow soil mixing (SSM) method, modiied for accommodation
of dry binders, utilises a crane mounted single auger tool 1.8–3.7 m in
 In-situ soil mixing 347

Table 9.1 Mixing conditions for selected dry deep mixing methods
Selected dry DM methods
Technical speciication DJM Nordic method Trevimix
Number of mixing 2 (standard), 1 1 2 (more common), 1
shafts
Diameter of mixing 1.0 (standard) 0.5–1.0 possible 0.8–1.0
tool [m] 1.3 (modiied) 0.6, 0.8 standard (standard)
Realistic maximum 33 25 (30) 30
penetration
depth [m]
Penetration/Retrieval 0.5–3 (4), 7 (1 shaft) P: 2–15 P: 0.4
velocity [m/min] typically: R: 2–6 R: 0.6
P: 1.5, R: 0.7, 0.9 (R: 15–30 mm/
(R: 15 mm/rev.) rev.)
Penetration/Retrieval P: 24, 32 (Electr.) R: 100–220 10–40
rotation speed R: 48, 64 (Electr.) (150–180 P: 20 typically
[rpm] P/R 21–64 (Hydr.) typically) R: 30 typically
Injection during R R R
Penetration/Retrieval (P used: air/binder) (P possible) (P used: air/binder)
Footprint area of the 0.78 : 1 × 1.0 m 0.28, 0.5 (0.78) 0.78 : 1 × 1.0 m
mixing tool (max.) 1.56 : 2 × 1.0 m 1.56 : 2 × 1.0 m
[m2] 2.65 : 2 × 1.3 m
Amount of injected 100–400 70–150 150–300
dry binder [kg/m3] Cem.:sands 150–250 organic 250 typically
200–600 Cem.: peat soils
50–300 Lime: clay
Binder supply 25–120 standard, 40–230 around 100
capacity per shaft up to 200 mod.
[kg/min] version
Injection pressure P: 100–600 400–800 600–1,000
[kPa] R: 600–100
Productivity [m3/shift] 300–700 150–300 150–220

diameter (Jasperse and Ryan, 1992). The driver for the tool is a drilling
system. It can be a conventional hydraulic drill or a high-torque dual motor
turntable. The auger tool itself is specially designed to break up the soil
and/or sludge and mix it with dry reagent without bringing the material
to the surface. To suppress emissions from the mixing process and/or for
environmental applications, the mixing tool can be enclosed in a hood or
bottom-opened cylinder to control dust and airborne contaminants (Figure
9.9). Further components of environmental control may also include a
low pressure blower or vacuum pump to keep negative pressure inside the
hood during operation, a dust collector, a fume incinerator, or an activated
carbon scrubber, depending on site-speciic conditions and contaminants
(Aldridge and Naguib, 1992).
348 Ground improvement

Figure 9.9 SSM mixing tool diameter 3.7 m for the dry method. (Courtesy of Geo-Con
Inc.)

Treatment reagents are transferred pneumatically to the mixing unit. The

delivery system consists of bulk storage tanks, several pneumatic pumps,
portable booster stations, and material receivers. The inal application of
the reagent to the treated soil is made with the hood lowered by dropping
the reagent into the hood through a calculated rotary valve located at the
bottom of a material receiver. Various cementitious, chemical, or even bio-
logical reagents can be added to soil or waste with this method.
The shallow mixing machines developed in Finland and Sweden are
essentially different from the column stabilisation machines. The mass
mixing tools are typically attached to the arm of a crawler mounted exca-
vator to enable vertical and horizontal movements of the tool through the
soil to complete mixing (Figure 9.10). The binder is fed from a separate unit

(a) (b)

Soft Soft Stabilised soil

Stabilised soil soil
soil

Figure 9.10 Mass stabilisation using dry binders. (a) Mainly vertical mixing. (b) Vertical
and horizontal mixing.
 In-situ soil mixing 349

which houses the pressurised binder container, compressor, air dryer, and
supply control unit. The operator injects the binder into the soil in such a
manner that it is equally distributed and mixed.
Different mixing tools have been tried in the past to treat very soft and
organic soils. The tools currently in use comprise cutting/mixing heads
equipped with blades rotating about a vertical axis, as shown in Figure
9.11, as well as very eficient cutting/mixing drums developed by Allu
Finland Ltd. (Powermixer systems PM and PMX) and suitable for large
mass stabilisation projects (Figure 9.12). The diameter of the mixing tool
rotating about a vertical axis is normally 600–800 mm, and the rotation
speed lies between 80–100 rpm. This method can be applied for soft clays
and organic soils with shear strength below 25 kPa. The more power-
ful PM/PMX mixing drums, with maximum torque of 7 and 23 kNm for
two drums, respectively, can be itted with different types of exchange-
able blades, teeth, or wings enabling enhanced mixing or cutting/break-
ing action in the treated soil. A typical mixing drum has plan width of
1.5–1.6 m and outside tooth to tooth diameter of 0.85–0.95 m. The work-
ing depth is 0–5 m. The novel aspect of the PMX system consists of drums
mounted on two inclined axes of rotation. This arrangement reduces the
distance between both drums and improves the homogeneity of mixing by
eliminating potential zones of unmixed soil which may appear behind the
mounting arm in case of the PM tools, especially when mixing soils with
higher shear resistance.
The mixing pattern of mass stabilisation is planned taking into account
site-speciic conditions and capabilities of the mixing machine and the

(a) (b)

Figure 9.11 Mass stabilisation of organic soil in Sweden. (a) Equipment. (b) Mass mixing
tool diameter 1.0 m. (Courtesy of LCM.)
350 Ground improvement

(a)

(b)

Figure 9.12 Mass mixing tools developed by Allu Finland Ltd. (a) Powermixer system PM
500, with cutting/mixing drums diameter 0.95 m, mounted on a horizon-
tal axis of rotation. (b) Powermixer system PMX 500, with cutting/mixing
drums diameter 0.85 m, mounted on two inclined axes of rotation.

mixing tool. Usual practice is to stabilise in one sequence a block of soil

within the operational range of the machine (e.g., 4 × 4 m in plan and 5
m deep). When the prescribed amount of binder is mixed into the volume
treated, remoulding is continued in order to obtain a homogeneous soil-
binder mixture. The productivity rate is approximately 200–400 m3/shift
 In-situ soil mixing 351

of stabilised soil, depending on the system/tool used and type of soil. In

heavier soils the production rate decreases. The amount of binder used is
typically in the range of 100–250 kg/m3. In Scandinavia the objective for
shear strength in peat is usually 50 kPa (Jelisic and Leppänen, 2003).

9.3.3 Wet method mechanical deep mixing

The wet DM methods applied for ground treatment on land in Japan, US,
and Europe are generally developed to produce similar quality columns
or panels/walls of stabilised soil, with unconined compressive strength
in the order of 0.5–5 MPa, or even more in granular soils, while the
machines, mixing tools, execution procedures and productivity differ
considerably.
Typical wet-method DM construction equipment consists of a batch
mixing plant to supply proprietary slurry, and of a mixing machine for
injection and mixing of slurry into the ground. The plant generally includes
silos, water tank, batching system, temporary storage tank, slurry pumps
(equipped with low metres), and power supply unit. Cutter soil mixing
and vertical trenching processes can be also supported by compressed air
and a desanding plant. The batching system can be varied from manual
or computer-controlled colloidal shear mixer to a very fast inline jet mix-
ing system. The storage tanks have paddle agitators to keep the compo-
nent materials from settling out of the slurry. Delivery pumps are duplex
or triplex reciprocating piston pumps, or variable speed progressive cavity
pumps. Pumping rates typically range from 0.08–0.4 m3/min, but can reach
up to 1 m3/min for high-capacity mixing tools. Any changes in the slurry
are made by adjusting the weight of each ingredient. Since luid volume is
being introduced into the ground, spoils must come to the surface.
The operational systems involving wet method mechanical deep mixing
can be subdivided into three groups, taking into account rotation of the
mixing tool about vertical or horizontal axes, or around the whole cutting
arm (cf. Figure 9.6).

9.3.3.1 Mixing tools rotating about vertical axes

The machines that are used for on-land applications usually have 1–4
shafts mounted on ixed or hanging leads, and are equipped with specially
designed mixing tools. A multi-axis gearbox distributes the torque from a
rotary drive unit to each shaft for penetration to the intended depth. The
penetration speed is typically in the range of 0.5–1.5 metres per minute
and is usually increased during withdrawal. The mixing tools are kept in
parallel by joint bands mounted at vertical intervals along the drive shafts.
With some machines the spacing between individual shafts can be adjusted
within prescribed limits to produce overlapped columns, which is beneicial
352 Ground improvement

when forming continuous panels or blocks of stabilised soil in a single-

stroke operation. A summary of mixing conditions for selected wet DM
methods used for on-land applications and utilising mechanical mixing is
presented in Table 9.2.
For marine applications, large execution vessels equipped with the mix-
ing machine, batching plant, storage tanks, and a control room are usually
used for rapid treatment of considerable soil volumes. The area of treatment
in single-stroke operation with 2–8 mixing shafts ranges from 1.5–9.5 m2,
and the productivity rates are in excess of 1000 m3 per day.
The mixing tools for the wet method are designed for various improve-
ment purposes and are conigured to soil type and available turning equip-
ment. Since there is no one tool that can successfully treats all soils, ield
adjustments are typical. The mixing tools can be broadly classiied into
blade-based and auger-based constructions (cf. Porbaha et al., 2001).
The tools of the irst group have an assortment of lighting and mixing
blades of full or near-full diameter and different orientation to eficiently
break down the soil structure. Steel hard-facing and an arrangement of
purposely located teeth serve to aid penetration and reduce maintenance.
A small diameter lead auger/drilling bit usually extends below the cut-
ting blades to centre and control penetration and verticality. The mixing
process is mainly conducted at or within a short distance from the tip of
the drill shaft(s). Injection nozzles are strategically located on the tool to
uniformly distribute the slurry into the soil. They are usually found near
the shaft tip, but can also be located along and above the mixing blades.
Sophisticated single-axis systems with double cutting/mixing blades,
spaced apart and rotating in opposite directions, have also been developed
in Japan (Horpibiulsuk et al., 2002) and in Germany (Bauer Machines,
2012). The counter-rotating components provide an exceptionally high
degree of shearing, especially in cohesive soils, and uniformity of soil-mix.
Examples of single- and multiple-shaft mixing tools are shown in Figures
9.13 and 9.15, and a single-column mixing tool for double rotary drives
is shown in Figure 9.14. The latter tool is designed in such a way that the
outer cutting frame, driven by a rotary head with a larger torque, is used to
loosen the soil whereas the faster rotating inner blades are used to enhance
the mixing process.
The second group employs discontinuous or continuous helical augers
for drilling and mixing or several levels of inclined paddles located above
the cutter head of the mixing shaft. In these systems interlocking or closely
spaced multiple shaft arrangements are typically used and the mixing oper-
ation is enhanced by counter-rotating action of adjacent shafts. The mixing
process occurs along all or a signiicant portion of the drill shaft(s). The
direction of rotation is usually reversed during withdrawal. In most sys-
tems the slurry is fed through nozzles located at the bottom of each shaft.
Examples of mixing tools are shown in Figures 9.16 through 9.18.
Table 9.2 Mixing conditions for selected wet DM methods used for on-land applications
Selected wet DM methods for mechanical mixing about vertical axis(es)
Technical CDM (Stand. CDM HB-Keller
speciication and MEGA) Land4 SCC USA/Europe Bauer SMW DSM COLMIX
Number of mixing 2 4 1 1, 2, 3 3 1-3, 5 1–6 usually 4 2, 3, 4
shafts 1:older syst. possible 2 possible 1 usually 3
Diameter of mixing 1.0: S 1.0/1.2 0.6-1.5 0.5-2.4 :U 3 × 0.37 0.55–1.5 0.8–1 usually 0.23–0.75 :2
tool [m] 1.2/1.3: M 1.2 (2 shafts) 0.6-1.2 :E 3 × 0.55 usually 0.9, 0.9 0.36–0.50 :3
(shaft spacing) (variable) (variable) 3 × 0.88 (variable) 0.50–0.75 :4
Realistic maximum 50 (55) 25 20 20 0.37: 10.5 35 (50) 35 20
penetration depth 30: M 1.2 0.55: 15.7
[m] 20: M 1.3 0.88: 25
Penetration/Retrieval P: (0.3) 0.5-1 P: 0.7–1 P: 1.0 P: 0.3–1 P: 0.2–1 P: 0.5–1 P: 0.6–1 P: 0.8
velocity [m/min] R: 0.7–1 (2) R: 1.0 R: 1.0 R: 1–2 R: 0.7–1 (5) R: 1.5–2 R: 1–2 R: 1.0
Penetration/ P: 20 P: 20 30–60 P: 20–25 20–40 (80) P: 14–20 15–25 8–30
Retrieval rotation R: 40 R: 40 R: 40–60 R: higher
speed [rpm]
Injection during P and/or R, P and/or R, usually P, P (+ R) :U P and/or R, P and R, P (+ R), P (+R)
Penetration and/ restroking at restroking restroking at P + R with P (30%–50%) restroking restroking at ev. restroking
or Retrieval the bottom at the the bottom restroking:E restroking common the bottom in clays
bottom

In-situ soil mixing

Water/Cement 0.6–1.3 0.6–1.3 0.6–0.8 clays, 1–1.5: U 0.6–2.5 0.7–2.5 1.2–1.75 0.7–2.5
ratio av. 1.0 av. 1.0 1.0–1.2 sands 0.6–1.2: E av. 1.5
Footprint area of 1.5: 2 shafts 2.83–3.14 0.3–1.75 usually 0.44: 3 × 0.37 0.7: 3 × .55 2.5 0.08–0.95 :2
the mixing tool 0.8: 1 shaft or 2.25 two 1.1–4.5: U 0.94: 3 × 0.55 1.7: 3 × 0.9 (4 shafts, 0.29–0.57 :3
[m2] 2.17/2.56: M 4.21–4.52 shafts 0.5–1.5: E 2.35: 3 × 0.88 4.7 :3 × 1.5 tangential) 0.76–1.60 :4
Amount of added 70–300 70–300 150–400 150–275: U 80–500 200–750 120–400 100–550
(dry) binder [kg/m3] av. 140–200 av. 140–200 250–450: E

353
Productivity per 100–200 m3 500–700 m3 100 m2 wall 250–750 m3: U 30–300 m3 100–200 m3 200–300 m2 100–300
shift (one rig) 400 m col. 75–200 m3: E wall linear m
354 Ground improvement

(a) (b)

Figure 9.13 Single-shaft mixing tools. (a) Diameter 0.8 m. (b) Diameter 2.4 m, with a
free-blade system. (Courtesy of Keller-Hayward Baker.)

Figure 9.14 Single-column mixing tool for double rotary drives (SCM-DH); column
diameter 1.8 to 2.4 m, max. depth 23.5 m. (From Bauer Maschinen GmbH.
(2010). Cutter Soil Mixing: Process and Equipment. Brochure No. 905.656.2.)
 In-situ soil mixing 355

(a) (b)

(c) (d)

Figure 9.15 Multiple shaft mixing tools. (a) Standard CDM 2 × 1.0 m. (From Cement
Deep Mixing Association. (2002). CDM Cement Deep Mixing Bulletin as of
2002.) (b) CDM Land4, 4 × 1.0 m. (From Cement Deep Mixing Association.
(2002). CDM Cement Deep Mixing Bulletin as of 2002.) (c) SMW mixing pad-
dles 3 × 1.5 m. (d) Cutter head. (Courtesy of R. Jakiel.)

In addition to the information presented in Table 9.2, another CDM

machine for on-land applications deserves more speciic attention. The
CDM-Column21 machine uses two shafts with large mixing heads of 1.5 m
(1.6 m) diameter consisting of an upper and lower mixing unit, both of
which are equipped with inner and outer mixing blades that rotate in
opposite directions (Figure 9.19). The unique counter-rotating action of the
blades accentuates the shearing mixing effect and ensures uniform mixing
356 Ground improvement

(a) (b)

Figure 9.16 DSM mixing tools. (a) Four blade-based mixing shafts. (b) Four discontinu-
ous, interlocking augers, diameter 0.9 m. (Courtesy of Geo-Con Inc.)

of the cement slurry with the soil. The unit is capable of treating harder
ground formations sandwiched between softer layers. The area of treatment
is 3.5 m2 and the required capacity of slurry supply is up to 1.0 m3/min.
The maximum depth of treatment is 40 m. This modern system not only
reduces the unit cost of soil treatment due to its very high productivity,

(a) (b)
Twin Triple

Quadruple
Direction of
rotation
during drilling

Figure 9.17 COLMIX mixing tools: (a) Possible arrangements. (b) Four discontinuous,
interlocking augers diameter 0.5 m. (Courtesy of Bachy-Soletanche.)
 In-situ soil mixing 357

(a) (b)

Figure 9.18 Bauer mixing tool with three closely spaced, continuous augers, diameter
0.55 m. (Courtesy of Bauer Spezialtiefbau.)

but also offers higher-quality soil improvement through increased mixing

operation eficiency (CDM Assoc., 2002).
Except for special situations and projects executed very close to sen-
sitive objects, wet method deep soil mixing has a very low impact on
nearby structures. To avoid net volume increase and corresponding
lateral stress in the ground caused by penetration of the mixing tool

Figure 9.19 CDM-Column 21 mixing tool. (Courtesy of CDM Association, Japan.)

358 Ground improvement

and injection of cement slurry, a dedicated method called CDM-LODIC

(low displacement and control) has been developed and modiied since
1985 (Sugiyama, 2002). In this system the upper part of the mixing
shafts are equipped with auger screws to forcibly expel equivalent soil
volume during penetration and withdrawal stages of the mixing tool
(see Figure 9.2). The screws have standardised dimensions (diameter
and pitch), and can be changed to best suit the ground conditions. It
has been demonstrated that the installation of CDM and CDM-LODIC
columns in soft clay 1.5 m from an inclinometer installed in a vertical
borehole causes at the depth of 17 m maximum horizontal displace-
ments of 16.11 and 1.01 cm, respectively, conirming eficiency of the
LODIC method (Horikiri et al., after CDIT, 2002). In addition to the
normal quality control system used for the conventional CDM method,
an automatic system has recently been developed to display the volume
of extracted soil. Since the cement slurry is normally injected during
withdrawal through the nozzles located above the mixing blades, the
extracted soil is free of cement and can be deposited or reused without
any restrictions, if not contaminated.

9.3.3.2 Mixing tools rotating about horizontal axes

The cutter soil mixing (CSM) method utilises special cutting and mixing
heads derived from diaphragm walling cutter technique. They are attached
to round or rectangular shaped Kelly bars, reaching penetrations of up to
about 20–40 m, respectively, or are suspended from ropes to reach depths
up to about 50 m. The standard head consists of two sets of counter-rotat-
ing, vertically mounted, cutting/mixing wheels, as shown in Figure 9.20a.
The wheels run on independent drives and can be driven in both directions.
They are equipped with cutting teeth capable of drilling and mixing even
in stiff ground and keying into bedrock. The in-situ soil is broken up, while
at the same time a speciic slurry is pumped to the nozzles centrally posi-
tioned between the wheels. The rotating wheels and cutting teeth push the
loosened soil through vertically mounted shear plates that have the effect
of a compulsory mixer to form homogeneous soil-cement panels, usually
2.4 or 2.8 m long and 0.55 to 1.2 m wide. Compressed air can be also used
to assist cutting and mixing operation during down stroke phase. Typical
penetration speed of the cutting/mixing head is 25–30 cm/min in sand and
gravel and 5 cm/min in cobbles. Withdrawal speed is usually 50 cm/min.
The CSM tool cuts vertical rectangular panels resulting in fewer vertical
joints when compared to multiple shaft systems and making this system
ideally suited for linear in-situ structures such as cut-off barriers, retaining
walls and liquefaction mitigation cells. Reinforcement in the form of steel
sections can be used to provide additional structural strength where needed.
An advantage of the method is that both fresh in fresh and fresh to hard
 In-situ soil mixing 359

(a) (b)

Figure 9.20 The cutter soil mixing heads. (a) Standard BCM 5, type 3-2 (three tooth hold-
ers in one row of teeth). (b) QuatroCutter. (Courtesy of Bauer Spezialtiefbau
GmbH.)

panel construction joints can be facilitated. For bigger depth, two standard
cutting and mixing heads attached at the bottom and at the top of a special
frame are used (Figure 9.20b). This system, known as QuatroCutter (Bauer
Maschinen, 2010), ensures intensive and homogeneous mixing as well as
high directional accuracy of CSM walls up to 60 m deep.
In relatively uniform soils, or for retaining structures up to about 20 m
deep, mixing is conducted during cutting (penetration) and withdrawal.
The backlow of soil and slurry is collected in a pre-excavated trench. For
deeper cut-off walls and for less uniform soils, a two-phase approach is
adopted for CSM using bentonite for temporary trench support during
cutting. As with conventional diaphragm wall construction, the bentonite
slurry is recirculated and cleaned by passing through desanding equipment.
Grout is injected during the withdrawal phase and mixed with remaining
soils. The speed of extraction and low of grout are adjusted to ensure that
the desired quantity of cement is blended with the soil.
The CSM system allows complete instrumentation inside the cutter gear-
box support frame to read and control in real time the coordinates of the
cutting head. This inclinometer system, coupled with the advantage of a
steerable tool, provides assurance of complete overlap between panels. In
favourable soil conditions the net productivity can reach about 40 m 2 of a
CSM wall/hour. However, a daily output is usually 100–200 m 2 (based on
70–90-tonne machines with power outputs of 260–300 kW) due to high
maintenance on the rig and cutter head. Wear rates are different for each
type of soil. For instance, in compacted sand and gravel re-welding of the
head was needed every 500–1,500 m 2 and the wear was 0.1–0.2 teeth/m 2 .
Changing of the head takes one day.
360 Ground improvement

9.3.3.3 Vertical trenching

A further facet of soil mixing has been provided by the development of
equipment which enables wet mixing while cutting trench structures in the
ground (e.g., FMI, TRD, and TRENCHMIX). Key advantages of these
single-phase walling methods, given the right conditions, are a reduction
in the number of joints over competing methods and less waste for any
required wall thickness as overlap is minimised, leading to reduced costs
and improved quality.
TRD equipment consists of a large machine about 100 tonnes and 7 m
tall and effectively enables cutting and mixing by means of a chainsaw
concept (Figure 9.3a and b). An initial starter trench is formed into which
a post is assembled and lowered into the ground to the required depth.
The post holds the cutting chain and injection proceeds as the machine
crawls forward cutting a full-depth face and providing a uniform mix in
place material devoid of the original soil stratiication. Joints in the soil-
mixed trench material only occur if production is stopped, for example if
only day shifts are being used. In this way joints are few and the system is
particularly eficient for long cut-off barriers (cf. Evans and Garbin, 2009).
Depth is limited to roughly 60 m. Wall thicknesses of 550–800 mm are
possible with presently available equipment. For earth retention applica-
tions, steel beams are inserted in the freshly constructed wall to provide the
required lateral strength. Productivity is highly inluenced by depth, width,
soils (rock), excavation support versus cut-off, and length of segments. For
trenches about 20 m deep and 700 mm wide, average productivity is about
250–300 m 2 /shift. The teeth can need changed once a week to once a shift
if working in hard soil/rock.
Quality control of TRD walls during construction includes monitoring
of the grout components and speciic gravity (SG) testing of the neat grout
in real time using a mass low sensor. The same sensor also measures and
records the low rate, volume, and temperature of the grout being pumped
through the system. It is typical to also verify SG several times each shift
using a mud balance as a check test for the instrumentation. Additionally,
the wet soil-mix material from the trench is sampled and subjected to low
table testing in order to assess the mix viscosity. Maintaining the low table
value within experimentally established range ensures proper material low
around the cutter post, which is essential for uniform full-depth mixing
(Garbin et al., 2010). Wall verticality is controlled by the operator and
monitored in real time using inclinometers installed inside the TRD cutter
post at various elevations. Additionally, the position of the cutter post can
be tracked using a differential GPS (Global Positioning System), as well as
with routine surveys using a total station.
TRENCHMIX uses a modiied Mastenbroek ditchdigger (Figure 9.21a).
The trench cutting chain is reversed to enable breaking up and mixing of
 In-situ soil mixing 361

(a)

(b)

Figure 9.21 TRENCHMIX equipment. (a) Rig in operation on a river dike. (b) Close-up
of cutting/mixing teeth. (Courtesy of Bachy Soletanche and Mastenbroek.)
362 Ground improvement

the soil with slurry injected at a controlled rate along the boom. Effective
and thorough mixing is ensured by the specially designed teeth
(Figure 9.21b) and high energy mixing process, controlled by a purposely
designed QC/QA system. The fact that the soil-mix material is drawn
to the surface allows also a good visual inspection of the eficiency of
mixing. Vertical mixed elements of about 0.4 m width to depths of typi-
cally 4–10 m (max 13 m) can be constructed. These elements can be used
to form cut-off walls to control pollution or groundwater in the ground,
or as improved ground foundation bearing elements when spaced close
together.
The trencher, despite its long boom, is highly manoeuvrable and capable
of operating in narrow and limited headroom spaces. This technique is
faster than sheet piling, produces signiicantly less spoil than convention-
ally dug walls and has no issues with noise and vibration. In favourable soil
conditions, the driving speed may reach 30–40 m/hour at the depth of 10
m and width of 0.4 m.
Mixing in the trench using dry binders is also possible given the right
conditions. The binder is placed in the shallow pre-trench and water is
added during mixing to achieve the required workability.

9.3.4 Wet method mechanical shallow mixing

Wet method mechanical shallow mixing can be used to improve substantial
areas of soft or loose supericial soils in ground engineering applications, as
well as for stabilisation and ixation of contaminated soils.
The SSM method uses specially designed single augers of 1.8–3.7 m
diameter, attached to a hollow-stemmed Kelly rod suspended from a crane.
Similar systems offer rigid attachment of the mixing tool to the base unit,
and can therefore incorporate down-pressure capability. The Kelly transfers
the torque and feed pressure to the mixing tool, while the swivel mounted
at the top of the rod seals the connection for delivery of the binder dur-
ing rotation. The binder is usually injected during penetration, in slurry
form, through several ports mounted at the bottom of the mixing augers.
The pitch on the auger lights and the centrifugal force help to distrib-
ute the binder to all parts of the column during rotation. Cycling up and
down with reduced binder delivery rates is often performed to improve
mixing eficiency. An overlapping pattern of primary and secondary col-
umns is normally used to ensure that the entire volume of treated soil is
thoroughly mixed. A high-torque driver in the range of 400–600 kNm
and high-capacity batching plant are generally required since the treatment
area may reach about 10 m3 of soil per meter of penetration. Examples of
large-diameter mixing tools are shown in Figure 9.22.
Wet method mass stabilisation can be also carried out with specially
designed mixing tools that are similar to those presented in Section 9.3.2.
 In-situ soil mixing 363

(a) (b)

Figure 9.22 Crane-mounted SSM tools for the wet method. (a) Diameter 2.4 m. (b)
Diameter 3.7 m. (Courtesy of Geo-Con Inc.)

Druss (2002) describes a major project conducted at the Fort Point

Channel site in Boston, where very soft organic sand and organic silt
deposits were shallow mixed prior to the execution of DM in underlying
marine blue clay. The works were mostly performed underwater, in areas
initially dredged to remove obstructions and timber piles. The objective
of stabilisation was to construct a temporary support for a drill bench
required for land-based DM operations. Shallow mixing was performed
using a sectional barge, an excavator with extended reach, and a shal-
low mixing bucket containing blades rotating about a horizontal axis
(Figure 9.23). Jet nozzles delivering the ly ash and cement grout were
located inside the bucket and were directed towards the mixing blades.
The bucket mixed horizontal trenches 1.2 m wide and about 10 m long in
1-m vertical lifts, moving from the surface of soft sediments to the top of
clay or inishing at partial depth.
A similar application has also been mentioned by Terashi (2002a). The
original ground was an artiicial landmass in Imari City, Japan, reclaimed
by dredged sea-bottom clay with undrained shear strength on the order
of 1 kPa. A 2-m thick block of treated soil was used to provide a working
platform and/or temporary access road loating on the extremely soft soil
deposit. In this particular case, a special loater equipped with four mix-
ing shafts was placed directly on the soft soil and dragged horizontally by
winches while the mixing tools were moved up and down vertically. Similar
shallow mixing tools as used in Finland and Sweden are also available in
Japan.
364 Ground improvement

(a)

(b) (c)

Figure 9.23 Shallow mixing equipment used at Fort Point Channel Site in Boston.
(Courtesy of R. Jakiel.)

9.3.5 Wet method hybrid deep mixing

In addition to mechanical mixing, these methods employ high-velocity
jet grouting in order to reduce penetration resistance and improve mixing
operation and/or to increase the diameter of the improved ground.
The SWING (spreadable wing) method, initially developed as a mechani-
cal mixing system, uses a retractable mixing blade mounted at the end
of a single drilling shaft. The position of the blade in the ground can be
changed from a vertical to horizontal alignment and vice versa, as shown
in Figure 9.24a and b. A combination of mechanical and jet mixing with
cement slurry enables columns of up to 3 m diameter to be constructed, and
the addition of compressed air allows columns greater than 3 m diameter.
During penetration of the ground, the soil is broken down by rotation of
 In-situ soil mixing 365

(a) (c)

(b)

Figure 9.24 Spreadable wing (SWING) method. (a) Blade position during penetra-
tion. (b) Blade expanded. (c) Demonstration of jetting action. (Courtesy of
SWING Assoc.)

the blade and jetting action of water. Cement slurry is injected during with-
drawal, with the jetting energy supplemented by air pressure. Air is used
when the larger diameters are required or when the soils under treatment
are too stiff. This method also enables the installation of inclined or even
horizontal columns and therefore allows soil mixing in areas of dificult
access.
The JACSMAN system consists of two 10-bladed mixing tools, each
combined with a pair of jet grouting nozzles aligned for Cross Jet (XJET)
to ensure that over-cutting does not occur. As compared with the con-
ventional CDM method, JACSMAN offers signiicant improvements that
contribute to a more economical, high-quality product. Due to XJET cut-
ting with air-enveloped, high-velocity cement slurry during withdrawal,
the treatment area of single-stroke operation increases considerably and
equals 6.4 m2 for type A arrangement, with a 75% share of jet mixing, and
7.2 m2 for type B arrangement, with a 63% share of jet mixing, as shown in
Figure 9.25. Moreover, the diameter of the soil-cement column can be con-
trolled and changed over the column’s length through stopping and starting
XJET action, not affecting the surrounding soil due to the dissipation of jet
energy at the cross point (Figure 9.26). This allows for the adjustment of
the column’s diameter to soil stratiication, as well as for controlled mix-
ing operations close to structures or excavation walls. The main operating
parameters of JACSMAN are as follows: jetting pressure 30 MPa, jetting
slurry low rate 4 × 150 l/min, air pressure 0.7 MPa, grout low rate 2 × 200
l/min, grout pressure 5 MPa, and withdrawal speed of 0.5 and 1 m/min
for type A and B arrangement, respectively (Kawanabe and Nozu, 2002).
366 Ground improvement

(b)
(a) Conventional CDM method
Cross jet injected section
1.0
Improved area : 1.5 m2
1.8 m
Jacsman
Type A Type B
3.3 m 3.7 m Mechanical section
1.0 0.65 1.3 0.5
2.3

Mechanical
1.0 1.4
section Single-rod installation
Improved area : 6.4 m2 Improved area : 7.2 m2

Figure 9.25 The JACSMAN method. (a) Comparison of treatment areas. (b) Exposed sin-
gle column. (Redrawn from Kawanabe, S. and Nozu, M. (2002). Combination
mixing method of jet grout and deep mixing, Proceedings from the Deep
Mixing Workshop 2002, Port and Airport Research Institute & Coastal
Development Institute of Technology, Tokyo, Japan.)

HYDRAMECH utilises mechanical mixing with a single shaft, itted

with 1.2-m diameter paddles and a 0.9-m diameter auger, in combination
with high-velocity grout injection at 40 MPa through eight 2-mm ‘hydra
nozzles’ on the outer edges of the mixing tool. HYDRAMECH is capable
of creating soil-cement columns with diameters of 2 m. Mechanical mixing

(a) (b)

Figure 9.26 JACSMAN mixing tool. (a) Twin head assembly (note grout nozzle in front
of the tool). (b) XJET demonstration with increased pressure. (Courtesy of
R. Essler.)
 In-situ soil mixing 367

occurs smoothly in the centre of the column, and chunks of material are
forced to the perimeter, where they are disaggregated by the jets. Treatment
with jets can be switched on and off throughout the column length to cre-
ate plugs of treated soil. Realistic maximum penetration depth is 20 m.
Penetration/retrieval velocity is 1–3 m/min, with 5–20 rpm during pen-
etration and 10–30 rpm during retrieval (additional mixing). Industrial
productivities are in the range of 250–500 m3/shift. The main objective
for developing this method was to improve on current jet grouting tech-
nologies that can create subsurface problems with the use of compressed
air. HYDRAMECH can create an extended diameter soil-cement column
without the injection of compressed air and still provide the continuous
overlap that is a very positive aspect of jet grouting systems, particularly
when installing horizontal barriers.
TURBOJET (GEOJET in the US) combines mechanical mixing with
single luid jet grouting technology. Jetting is used during insertion of the
tool to increase penetration velocity while extraction is carried out solely
with mechanical mixing. A specially designed mixing tool (or processor),
itted at the end of a tubular Kelly bar, consists of two levels of inclined
blades and is furnished with several 4–8 mm diameter high-pressure nozzles
mounted along the shaft and tip (Figure 9.27). The exact nature and com-
position of the processor can be varied, depending on soil conditions. Grout
can be pumped with a discharge rate of 450 l/min at 30 MPa, although lower
lows rates and pressure (15 MPa) are the norm. Tool diameters range from
0.6–1.5 m, usually 0.9–1.2 m, and the practical available depth of treat-
ment is 25 m (Lebon, 2002). Instantaneous rates include 2–12 m/min (6 m/
min typical) during penetration and 15 m/min during withdrawal. Computer
control of the equipment during column formation is therefore required. The
computer analyses the rate of tool rotation and penetration, slurry pressure,
torque, crowd force, and soil-mix volume and density as a function of depth.
The system also reacts to changing parameters and automatically adjusts to
maintain speciied soil-cement properties, even in varying subsurface soils.
Because of the additional mixing energy supplied, restroking is not required.
Industrial production rates in excess of 150 m/h and 1100 m/shift are possi-
ble. The system produces low waste volumes (20%–30% of ground treated).

9.3.6 Installation process (mixing about vertical axis)

The typical installation process consists of positioning the mixing shaft(s)
above the planned location, penetration of the mixing tool, veriication
and improvement of the bottom soil layer, withdrawal, and movement to
a new location if necessary. The details of execution depend on the type of
method applied (dry or wet), technical features of the equipment, and the
site-speciic and functional requirements. Frequently used execution proce-
dures are shown schematically in Figure 9.28.
368 Ground improvement

(a) (b)

Figure 9.27 TURBOJET deep soil mixing equipment and processor. (Courtesy of M.
Siepi.)

(a) Time (b) Time (c) Time

Depth Depth Depth

(d) Time (e) Time

Depth Depth

Figure 9.28 Typical execution procedures of deep soil mixing. (a) Without or with bot-
toming. (b) With reversal(s) during penetration. (c) With bottom restrok-
ing. (d) With stepped restroking during withdrawal. (e) With full-depth
restroking.
 In-situ soil mixing 369

The position and verticality of the shaft is checked irst, and zero adjust-
ments of the logging system are conducted. For on-land application optical
survey devices are normally used, whereas for marine operations the use of
the GPS has become common. The GPS is also advantageous in the case of
large on-land projects, especially those involving treatment of very weak
supericial soils.
During penetration the mixing tool is delivered to the required depth. In
this phase compressed air (dry method), or slurry (wet mechanical mixing),
or high-velocity jetting with slurry or water and air (hybrid mixing) is used
to support mechanical drilling. Mechanical penetration may be dificult
when the tool hits a hard layer or when the improvement depth is relatively
deep, leading to possible damage or deadlock of the tool in the ground.
This danger may be reduced with partial restroking to minimise rotation
resistance along the shaft (Figure 9.28b), or by means of pre-boring with
an auger machine.
After the basal treatment depth is reached, the mixing tool remains
on the bottom rotating about 0.5–2 minutes for complete mixing. This
phase is often called ‘bottoming’ and serves to ensure suficient contact of
the column(s) with the bearing subsoil. Penetration into the bearing layer
should be conirmed by a rapid change of penetration velocity of the tool,
required torque, and rotation speed. At this stage the tool can be raised
about 0.5–1 m and lowered again to treat more effectively the transition
zone between soft and bearing soils (Figure 9.28c). Withdrawal may be
conducted as a continuous upstroke, but can include stepped or even full
restroking if needed (Figure 9.28d and e). Full restroking is beneicial in the
case of interchanged soft/stiff layers and stratiied soils, leading to more
uniform properties of stabilised soil across the depth of treatment.
The accompanying delivery of the stabilising agent to the subsoil is oper-
ator/computer controlled and linked to the energy of mixing in the speciic
layers of treated soil. In general, injection of the stabilising agent can take
place during penetration, withdrawal, and restroking; however, two main
injection methods are distinguished—the penetration injection method,
which is a top-bottom process, and the withdrawal injection method,
which is a bottom-top process. Penetration injection is typically used for
on-land applications of the wet method because the slurry helps to lubri-
cate the mixing tool and assists in breaking up the soil into smaller pieces.
Normally, 80%–100% of the total slurry volume is used in this stage. This
method is also beneicial to the homogeneity and strength of the manufac-
tured column because the native soil is mixed twice with the binder.
Withdrawal injection is typically used for the dry method, usually
with the whole amount of binder delivered to the soil during this phase.
However, if very high binder concentration is needed to reach the design
strength, part of the stabiliser may be injected during penetration phase
and the rest during withdrawal of the mixing tool. Withdrawal injection
370 Ground improvement

also prevails for marine operations with the CDM method, but not for all.
For on-land applications with the wet method, withdrawal injection is also
possible, but usually at a reduced low rate to minimise the volume of spoil
(except of the CDM-LODIC method where withdrawal injection is a stan-
dard installation process).
The sequence of mixing operations will need to be adjusted to suit each
site’s speciic conditions, but in general the most eficient sequence is to
work the stabilisation machine within its radius of operation as much as
possible before it is moved.

9.3.7 Details of construction and execution

(mixing about vertical axis)
9.3.7.1 Number of shafts
Models and ield observations indicate that multiple-shaft arrangements
generally provide better homogeneity of DM columns than those pro-
duced with single-shaft mixing tools furnished with ixed cutting/mix-
ing blades rotating in one direction. This has been especially observed in
clays, which may tend to stick to the mixing blades and hence rotate with
the mixing shaft, resulting in poor mixing. As a countermeasure, nonro-
tating vanes have been mounted on single auger shafts, close to cutting
blades. An example is the free blade, developed in Japan (CEDIT, 2002),
which extends beyond the reach of the mixing blades and is therefore sup-
posed to stay in the ground to provide sharing capability (cf. Figure 9.13b).
The ‘entrained rotation’ phenomenon is signiicantly reduced with closely
spaced augers when the neighbouring shafts rotate in opposite directions,
or are eliminated with overlapping augers or interlocking mixing blades
providing greater soil shearing and particle milling (the same applies for
sophisticated single-shaft tools with counter-rotating blades). Transverse
steel bars used to keep multiple shafts in position have a similar function
as the free blade.
Besides contributing to interactive mixing and increased productivity
rates, multiple shaft arrangements also minimise the countermovement
against shaft rotation as compared with single-shaft rigs, and improve sta-
bility of the machine. This further contributes to more precise control of
shaft verticality during penetration. Linear arrangements of mixing shafts,
applied for the construction of retaining and cut-off walls, enable easier
and safer connection of individual wall panels using the intercut principle.
Furthermore, limited adjustments of deviations occurring during penetra-
tion can be made by altering the rotation of coupled shafts.
Multiple shaft arrangements are, however, more demanding in terms
of constructional requirements, generally leading to more complicated
mechanical systems and larger/heavier machines. This may result in
 In-situ soil mixing 371

reduced lexibility in some applications, as well as increased mobilisation

and operational costs, as compared with single-shaft machines.

9.3.7.2 Shape and orientation of mixing blades

The function of the mixing tool is to disaggregate the soil during penetra-
tion and to facilitate binder injection and immixing with the native soil.
In the case of purely mechanical interaction, the mixing tool should also
create the appropriate column diameter. A wide variety of different mixing
tools have been tried so far, ranging from very simple to quite complicated
constructions, with the obvious outcome that no single construction method
can equally serve all soils. Nevertheless, some general indications can be for-
mulated, keeping in mind that when compared to the wet method, the dry
method requires more vigorous mixing to achieve the same level of homoge-
neity of soil-binder mix.
During downward movement of the shaft the mixing tool has to loosen
the soil, while during withdrawal the soil should be thoroughly mixed with
the binder and recompacted as much as possible to reduce excess spoil and
to ensure maximum mixed soil density (recompaction does not matter in
saturated conditions but is rather important for the dry method). This
generally occurs when the inclination of the blades to the rotating direc-
tion produces mixing movements from the outside inward and from above
downwards, opposite to the lifting movement of the tool. The degree of
mixing increases when the soil is inely divided into horizontal, inclined,
and vertical directions during tool rotation. This explains why in the case
of single-axis shafts the window-type mixing tools, such as shown in
Figure 9.8a and b, may perform slightly better than the tools with sev-
eral separated horizontal blades, as also corroborated by the investigation
conducted by Abe et al. (CDIT, 2002). On the other hand, soils like peat
require more shearing action to be thoroughly mixed, and this is usually
better achieved with multiple horizontal blades. Furthermore, window-
type tools cannot overlap when used in multi-shaft arrangements. In the
case of continuous or discontinuous augers, counter-rotation against the
direction of auger pitch permits the soil-binder mix to be recompacted dur-
ing withdrawal. Counter-rotation and/or shear bars also generally reduce
the mixing energy required.
When designing a mixing tool, it is usually necessary to ind a balance
between a good ability to penetrate stiff or compacted layers of soil and a
good mixing performance in soft soils. The same applies to the speed of
rotation and associated higher wear of the mixing blades, as well as the
possibility of easy repair and quick replacement of the mixing equipment.
Consequently, the goal of designing a mixing device that creates suficient
movement in the soil without a great mixing effort or long mixing time is
dificult to achieve.
372 Ground improvement

9.3.7.3 Position of injection nozzles

To ensure optimum mixing eficiency, the position of injection nozzle(s) is
different for various methods and installation processes. For the penetra-
tion injection method the outlet is normally placed close to the bottom end
of the mixing tool, while for the withdrawal injection method it is above
the mixing blades or at the level of the upper blade. Besides these two stan-
dard outits some mixing tools have nozzles at both levels to also facilitate
a combined penetration/withdrawal injection (e.g., CDM and DJM meth-
ods). During the penetration phase, the lower port is used and the upper
one is closed. When withdrawing, the opposite combination is applied.
With the wet method and single-axis mixing tools there is usually one centre
injection nozzle located close to the shaft tip, while large-diameter tools have
several injection nozzles located along the blades at speciied distances from the
central shaft. In multiple-axis tools the grout is usually fed independently to
each shaft, with the outlet port placed at the shaft tip. In some linear arrange-
ments grout can be also supplied through the central shaft, incorporating 1
or 2 nozzles at the bottom, depending on the auger diameter, as done for the
Bauer triple method. The direction of grout injection is generally horizontal.
As for the mixing tools of the hybrid method, the high-velocity jet noz-
zles are purposely located on the outer ends of the mixing blades to increase
the range of mixing, but they can be also located at the tip or along the
Kelly bar if jetting is primarily used to increase the rate of mixing tool
penetration. The direction of the jet stream may be horizontal or inclined,
depending on the nozzle orientation.

9.3.7.4 Degree of mixing

The eficiency of in-situ soil mixing with a stabilising agent is one of the key
factors affecting column homogeneity and strength. The degree of mixing
depends on the mixing time, type of mixer, characteristics of the native
soil, and the form of applied binder (slurry or powder) and the energy of
injection (low or high output velocity). The overall mixing process is rather
complex, especially for the dry method (cf. Larsson, 1999), and dificult
to quantify. Therefore, in an attempt to specify a criterion for the required
mixing work, which could be controlled and altered on site during execu-
tion, a simpliied index named ‘blade rotation number’ has been introduced
in Japan (e.g., CDIT, 2002). The blade rotation number, T, is deined as the
total number of mixing blades passing during 1 m of single shaft movement
through the soil, and is expressed as follows, considering:

(a) complete injection during penetration and outlet located below the
blades:
T = ΣM × (Rp / Vp + Rw / Vw ), (9.1)
 In-situ soil mixing 373

(b) complete injection during withdrawal and outlet located above the
blades:
T = ΣM × (Rw / Vw ), and (9.2)

(c) partial injection during penetration and main injection during with-
drawal, with the lower outlet active only during penetration and the
upper outlet active during withdrawal:

T = ΣM × (Rp / Vp × Wp / W + Rw / Vw ) (9.3)

where: T = blade rotation number [rev/m], ΣM = total number of mixing

blades, Rp = rotational speed of the mixing tool during penetration [rev/
min], Vp = penetration velocity [m/min], Rw = rotational speed of the mix-
ing tool during withdrawal [rev/min], Vw = withdrawal velocity [m/min],
Wp = amount of binder injected during penetration [kg/m3], W = total
amount of injected binder [kg/m3].

The total number of mixing blades, ΣM, is assessed by counting all cut-
ting/mixing blades that are effective in the mixing process, taking into con-
sideration the method of injection and position of the injection outlet(s) in
relation to the blades. A full-diameter blade is counted as two blades. For
example, when the outlet port is located beneath two levels of blades and
when injection is carried out during penetration, as is common for the wet
methods, the total number of mixing blades is ΣM = 4 and Equation 9.1
is used to evaluate T. In case of the withdrawal injection method and the
outlet port located above all blades, as is common for the dry methods, ΣM
is also four but only the withdrawal stage is considered (Equation 9.2). This
example demonstrates that higher rotational speeds are required for the
withdrawal injection method to obtain a comparable degree of mixing. In
case multiple restroking along the whole depth is used, as shown in Figure.
9.27e, the resulting blade rotation number is a sum of T values calculated
for each penetration/withdrawal cycle of the mixing tool.
The blade rotation number is used for mechanical mixing only, and the
soil conditions are included indirectly (i.e., through selection of appropri-
ate input values), taking into account accumulated experience and techni-
cal speciications of the equipment. Based on ield data obtained in loose
sands (Mizuno et al.) and clays, the blade rotation number of 360 has been
recommended in Japan for the wet method to ensure reasonably low value
of the coeficient of variation, ν, of the unconined compressive strength
(CDIT, 2002). Field tests using wet mixing in silty/sandy clay in Poland
revealed T = 430 to satisfy ν ≤ 0.3 (Topolnicki, 2009). For the dry mixing
methods the blade rotation number is typically 274 or 284 for the DJM
and 200 to 400 for the Nordic method, noting that dry binders are injected
mainly only during withdrawal. For special mixing tools using cutting/mixing
374 Ground improvement

blades that rotate in opposite directions on a single shaft, like for instance
the Bauer SCM-DH and the CDM Column 21 methods, there is a need to
conceive a new guideline for the quality of mixing.

9.3.7.5 Control of binder supply

The amount of binder injected in a certain soil volume is easier to control
for the wet method than for the dry method, where the binder is fed into a
stream of compressed air.
The wet mixing process blends the materials with water to form a slurry
at the design water to binder ratio. The quantity of binder components
needed for each batch is weighed and added to the measured water volume
in the mixer. Alternatively, ready-batched or preweighted bagged materials
can be used to simplify this process. The binder slurry is then transferred
to a temporary storage tank that continually agitates the slurry to ensure
that the constituents of the mix do not separate. The slurry is then pumped
at a speciied low rate to the mixing tool. In order to obtain the required
amount of binder per soil volume, the penetration and withdrawal veloci-
ties of the mixing tool and the applied low rates have to be simultaneously
adjusted, taking into account the number of restroking passes with slurry
injection. The low rate of binder slurry is controlled at the delivery pump
and monitored with a low meter.
With the dry method, the weight loss of the binder storage tank, con-
tinuously measured by means of load transducers and averaged in such
a manner that acceleration components are cancelled out, is used as an
indicator of the amount of used binder. This information is combined with
the corresponding geometry of stabilised soil to evaluate the binder output
in kg/m of column, or in kg/m3 of treated soil. To reach the predetermined
target value it is necessary to control the feeding rate of binder into the air
stream until the speciied rate of loss of the material is obtained. This is
mainly accomplished by adjusting the rotation speed of an impeller pro-
vided at the bottom outlet of the binder storage tank (Figure 9.29). The
feeding mechanism must be manufactured with a very high precision since
the distance between the rotating blades and the cylinder walls is on the
order of 1/100 mm. However, the throughput of the impeller is not a linear
function of the rotating speed and also depends on: (1) wear of impeller,
blades, and wall; (2) pressure and amount of binder in the tank; (3) air
and material low below the impeller; and (4) low properties of the binder,
making binder output control more sophisticated (Bredenberg, 1999). The
downward movement of binder in the tank towards the impeller is facili-
tated by ‘luidisation’ of the stored material, caused by blowing compressed
air from the tank bottom. To ensure high productivity with this system, it
is important that the air blown into the binder storage tank is suficiently
dry and that the binder material is free of particles able to cause blocking
 In-situ soil mixing 375

(a) (b)
Pressurised tank Feeding wheel

Main
air
Air for
Binder
rubber
valve
Bypass
Air
air
Cell feeder
Rubber Outlet
To mixing tool valve

Figure 9.29 Binder feeding systems. (a) Cell feeder used in the Nordic method (cour-
tesy of LCM). (b) Impeller used in the DJM method. (Redrawn from Aoi,
M. (2002). Execution procedure of Japanese dry method (DJM), Proceedings
of the Deep Mixing Workshop 2002, Port and Airport Research Institute &
Coastal Development Institute of Technology, Tokyo, Japan.)

or damage to the feeding control system. Moreover, suitable hose diameter

for the equipment used must be carefully selected to ensure smooth binder
low to the mixing tool.

9.3.7.6 Control during construction

Soil mixing, like other ground improvement technologies, uses indirect
control measures to ensure the quality of work and product during execu-
tion. The main objective of a control system is to ensure delivery of a cor-
rect amount of binder and mixing energy along the installed element. The
extent of in-situ mixing operation monitoring is closely associated with the
type of project and the required level of quality control.
For a typical production process of DM, the following should be docu-
mented: element identiication and/or position, mixing tool details, work-
ing grade, mixing depth, start time, time at bottom, inish time, mixing
duration, agent speciication, injection low rate and pressure, total amount
of agent used, tool rpm on penetration, tool rpm on withdrawal, and torque
of the shaft. From this information the mixing energy and binder content
can be calculated to match laboratory and/or test columns. The standard
criteria to ensure quality of tip bearing are the penetration velocity and the
applied torque. Centralised control systems are usually available to digi-
tally record all parameters and display information at the control panel to
facilitate real-time adjustments (e.g., Yano et al., 1996, Bredenberg, 1999,
Burke, 2002, Hioki, 2002). They also simplify the task of preparing daily
reports by recording the daily performed activities of soil mixing works.
376 Ground improvement

In applications requiring automatic and/or more sophisticated control, a

variety of measuring systems can be used to control the mixing process or
column verticality or to observe horizontal and vertical ground displace-
ments. The computer reacts to changing ground conditions and automati-
cally adjusts injection output to ensure speciic treated soil parameters are
provided for each stratum.

9.4 APPLICATIONS AND LIMITATIONS

9.4.1 Areas of application

The main areas of SM applications are as follows:

(1) Foundation support

(2) Retention systems
(3) Ground treatment
(4) Liquefaction mitigation
(5) Hydraulic cut-off walls
(6) Environmental remediation

Case histories relating to each application group may be found in the

cited bibliography as well as in the proceedings of specialty international
conferences on deep mixing, held in Tokyo (1996), Stockholm (1999),
Helsinki (2000), New Orleans (2003, 2012), Stockholm (2005) and Osaka
(2009). It should be pointed out, however, that in many cases SM works are
conducted to fulil combined functions. Consequently, certain projects fall
into more than one general category of application.
Selected case histories are included in Section 9.7 and in Chapter 10.

9.4.2 Patterns of deep soil mixing installations

Soil mixing can be done to a replacement ratio of 100% wherein all the soil
inside a particular block is treated, as is usually the case for shallow mix-
ing applications, or to a selected lower ratio, which is often practised with
deep mixing. The chosen ratio relects, of course, the mechanical capabili-
ties and characteristics of the applied method. Depending on the purpose
of deep mixing works, speciic conditions of the site, stability calculations
and costs of treatment, different patterns of column installations are used
to achieve the desired result by utilising spaced or overlapping and single or
combined columns. Typical patterns are presented in Figure 9.30.
Square or triangular grid patterns of single or combined columns are
usually applied when the purpose of SM is reduction of settlement and,
in some cases, improvement of stability. Common examples are road and
railway embankments. Walls are used for excavation control, to stabilise
 In-situ soil mixing 377

(a) (b) (c) (d) (e) (f)

(g) (h) (i) (j)

(k) (l) (m) (n)

Figure 9.30 Examples of deep soil mixing patterns. (a), (b) Column-type (square and
triangular arrangement). (c) Tangent wall. (d) Overlapped wall. (e) Trench/
CSM wall. (f) Tangent walls. (g) Tangent grid. (h) Overlapped wall with but-
tresses. (i) Tangent cells. (j) Ring. (k) Lattice. (l) Group columns. (m) Multiple
trenches/CSM walls. (n) Block.

open cuts and protect structures with shallow foundations surrounding the
excavation, and as a measure against seepage. They are also constructed to
increase the bearing capacity of improved soil against horizontal or sliding
forces, with column rows installed in the direction of horizontal loading or
perpendicular to the expected surface of failure. Walls can be constructed
with tangential or overlapping columns, joint panels, or as trench struc-
tures. Overlapping is particularly important when executing cut-off walls
or environmental barriers. In the case of DM machines equipped with
linearly arranged multiple shafts, walls are usually executed using inter-
secting primary and secondary panels, with partial or even full-column
diameter overlap. Groups of columns can be utilised to support embank-
ments and foundations in order to reduce settlements and/or increase the
bearing capacity. Various combinations of columns or panels are also used
to build grid, U-formed, cellular, or circular installations with tangential
or overlapping elements to improve the interaction with the untreated soil.
Lattice-type improvements are considered an intermediate, cost-effective
system between the wall-type and the block-type improvement. Full blocks
are used to create large, highly stable volumes of stabilised soil, which act
as gravity structures.
 Next Page
378 Ground improvement

To compare various patterns in terms of the treatment area and to eval-

uate composite properties of the treated elements and the surrounding
untreated soil, a purposely deined ratio of area improvement, ap, is used
(cf. Figure 9.31):

At net area of soil mixing

ap " " . (9.4)
A respective to
otal area

The upper limit of the ratio of area improvement for a square grid of tan-
gential columns is 78.5%, and for equilateral triangular grid it is 90.7%.
For columns spaced at 2 diameters ap is 19.6% and 22.7% for square and
triangular patterns, respectively; and for columns spaced at 3 diameters it
is 8.7% and 10.1%. The spacing of 3 diameters, usually recommended to
minimise interaction between piles, can be considered as a practical lower
limit of the area improvement ratio. Numerous embankments in Japan
have been stabilised with ap usually 30%–50% (due to seismic excitations),
while in Scandinavia area ratios 10%–30% have been typically applied in
case histories. Statistical evaluation of about 2,770 embankment projects
in Japan revealed a distinct difference between ap values used for settlement
reduction and stability problems, being about 20% higher for the latter
cases (Terashi et al., 2009).
Column/panel installation patterns may not only vary in plan view but
also with respect to the depth of treatment. In the wall-type improvement,
short and long walls can be alternately installed in the soft soil to reduce
the costs of soil mixing (Figure 9.32a). The long walls transfer the loads
exerted by the superstructure and external excitations to the bearing stra-
tum, while the intermediate short walls provide connection between the
long walls, increasing the rigidity of the total improved soil mass. This
type of improvement has been commonly applied in port and harbour con-
structions in Japan (e.g., Kansai Airport man-made island; CDIT, 2002).
Another example is the variation of column/panel lengths in transition
and/or purposely determined zones of soil treatment, as shown in Figure

(a) L1 (b) L

L2 B
Ac Aci
A

A = L1 × L2 , At = Ac A = B × L , At = ¦ Aci

Figure 9.31 Evaluation of the ratio of area improvement. (a) Regular grid of columns. (b)
Foundation slab.
Chapter 10

Dry soil mixing

Marcus Dahlström

CONTENTS

10.1 Introduction ................................................................................. 436

10.2 The dry soil mixing process.......................................................... 437
10.2.1 General .............................................................................. 437
10.2.2 The mixing process............................................................ 438
10.2.2.1 Phase 1: Penetration of the mixing tool .............. 439
10.2.2.2 Phase 2: In-situ mixing and dispersion of binder.... 440
10.2.2.3 Phase 3: Molecular diffusion after installation ... 441
10.2.3 Factors important in the mixing process............................ 441
10.2.3.1 Rheological soil properties ................................. 442
10.2.3.2 Type and amount of binder................................. 442
10.2.3.3 Mixing energy .................................................... 443
10.2.3.4 Mixing tool design and drilling rig site–
speciic adjustments ............................................ 445
10.2.4 Binders and soil properties................................................. 445
10.2.4.1 Binders................................................................ 445
10.2.4.2 Applicable soil types ........................................... 446
10.2.4.3 Binders and reaction process............................... 446
10.2.4.4 Strength gain ...................................................... 448
10.2.4.5 Long-term strength............................................. 452
10.3 Equipment, monitoring, and control............................................. 452
10.3.1 Equipment ......................................................................... 452
10.3.2 Monitoring and control ..................................................... 455
10.4 Application, design and testing..................................................... 456
10.4.1 Applications....................................................................... 456
10.4.1.1 Reduction of settlements..................................... 457
10.4.1.2 Improving stability ............................................. 457
10.4.1.3 Reduction of vibrations ...................................... 459
10.4.1.4 Foundations of structures and houses................. 459
10.4.1.5 Mass mixing and stabilisation of highly
organic soils and dredged muds .......................... 461
10.4.1.6 Solidiication and stabilisation of
contaminated soil ............................................... 463
10.4.2 Design................................................................................ 463
10.4.2.1 Design models..................................................... 464
10.4.3 Testing and quality control ................................................ 472
10.4.3.1 Field test methods ............................................... 472

435
436 Ground improvement

10.4.3.2 Performance of penetration tests......................... 478

10.5 Case histories................................................................................ 480
10.5.1 Thames Estuary, UK........................................................... 480
10.5.2 Jewish Creek, Key Largo, Florida ...................................... 482
10.5.3 Road 45 and Norway/Väner Route, Sweden ...................... 485
10.5.4 Railway embankment in Malaysia ..................................... 488
References ............................................................................................. 490

10.1 INTRODUCTION

Dry soil mixing (DSM) methods are the only techniques which directly
mix soils with dry binder materials where the soil moisture is suficient
to hydrate the resulting in-situ mix. A range of binders can be used, but
the most common are cement, lime-cement, and other cementitious blends
which undergo beneicial chemical reactions with the soils into which they
are mixed. These dry materials are fed into the ground using compressed
air where they are comprehensively mixed, using purpose-designed tools,
with the soils to be improved to the depth range identiied.
Early development of in-situ dry mixing methods appears to have arisen
simultaneously in Japan and Sweden during the late 1960s in order to
deal with soft silts and clays. Since then the approaches have markedly
diverged to address the differing technical and commercial demands of
their respective markets. This chapter deals speciically with the Swedish
method, sometimes dubbed the ‘Nordic method’ (Holm, 2002). By the end
of the 1980s the method was widely established in Sweden, Finland, and
Norway, and by the early 2000s was being used in the US and in several
other European countries, notably Poland and the UK. At the time of writ-
ing the method is also being used in the Far East and Australia. The his-
tory and development of both Nordic and Japanese methods are covered in
more detail in Chapter 9 (Section 9.2) of this book.
Single-axis column mixing is the simplest and most widely used form
of construction using the Nordic method. The mixing tool is irst rotated
into the ground down to the depth previously identiied for the toe of
the column. The tool is then withdrawn at a high rate of rotation, dur-
ing which the dry binder is injected through ports in the mixing tool at a
controlled rate according to the design. This creates appropriate mixing
of the binder with the soil contained within the blade diameter and shape
of the tool. Typical equipment for column construction is a purpose-built
hydraulic base machine with low ground pressure, suitable for traversing
soft ground, carrying a mast, rotary head, hollow-stem Kelly bar and the
mixing tool. A separate pressurised container is used to store and transfer
the binder to the Kelly, and is either self-propelled or towed behind the
base machine.
 Dry soil mixing 437

Dry soil mixing is applied in inorganic soft soils in which the natural
moisture content is close to or above the liquid limit, and is increasingly
being used in highly organic soils and peats. The main applications are for
foundation bearing or settlement control and providing stability for slopes,
embankments, and excavations. Other applications include reduction of
vibrations, provision of liquefaction resistance, and solidiication of con-
taminated soils and mud. The technique has continued to spread worldwide
due to its advantages of avoiding the need to pre-mix materials with water
before injection and the high rate of construction, leading to low costs per
metre of column relative to other forms of soil mixing. Typical column
diameters are 0.5–1.0 m and depths of treatment are often 5–17 m with
deeper treatment available with special equipment.

10.2 THE DRY SOIL MIXING PROCESS

10.2.1 General
Dry soil mixing is a general term for mechanical in-situ mixing of soil by
adding a dry additive (commonly called binder) distributed by air low. The
mechanical mixing is done by either vertical or horizontal mixing by rotat-
ing impellers of paddles see Figure 10.1 or by cylinders with cutting heads
as in Figure 10.2.
The binder is transported from rig-mounted silos or from silos mounted
on a separate unit (which could be either on a so-called shuttle or distrib-
uted directly from a bulk silo). The DSM process with the in-situ mixing of
a dry additive into the soil by compressed air is referred to by several differ-
ent terms, such as lime/cement columns, deep stabilisation, dry jet mixing

Figure 10.1 Vertical mixing tool (Pinnborr). (Courtesy of LCM.)

438 Ground improvement

Figure 10.2 Horizontal mixing tool (Allu). (Courtesy of Allu.)

methods, dry deep mixing, column stabilisation, or mass mixing. These

terms are explained in a later section.
The use of compressed air as the medium for transporting the binder is
advantageous because it takes a relatively small amount of binding agent
to achieve the requisite strength gain in the soil. Hence, soft soils contain
large amounts of water, and by adding a dry binder instead of wet slurry (as
is necessary in wet mixing or jet grouting processes) less binder is required.

10.2.2 The mixing process

The purpose of the mixing process is to distribute the binder into the soil
eficiently in order to provide the best possible condition for the chemical
reaction to take place. Depending on the depth and volume of the mixed soil,
the mixing process is most frequently applied using columns of mixed mate-
rial. Columns are typically 0.5–1.0 m in diameter and range 3–25 m in depth
and are solely installed vertically (minor inclinations at 4:1 to 10:1 exists).
Alternatively, for shallower depths, mass mixing may be appropriate.
Mass mixing can either be installed as interlocking columns into a block of
vertical elements or as a mass-mixed soil volume with a horizontally rotat-
ing cylinder with cutter heads, see Figure 10.2. Mass mixing is generally
performed in depth of 0.5–6 m; however, mass mixing using the interlock-
ing column method has been performed up to 15 m.
The production process for DSM is similar for columns and mass mixing
and can be divided into three principal phases:

(1) Penetration of the mixing tool to required depth

(2) In-situ mixing and dispersion of binder
(3) Molecular diffusion
 Dry soil mixing 439

The installation process

1. Drilling down 2. Injection of 3. Hardening 4. Embankment

binder with an placement
inverse mixing
tool rotation

Figure 10.3 Installation sequence for column installation.

The installation of columns is presented in Figure 10.3, and the installation

process of mass mixing is presented in Figure 10.4. For the mass mixing
method with interlocking columns the installation sequence of columns is
represented but with minor modiications.
In the European Standard EN 14679 2005 igure A.5, the installation
sequence is described as a ive step procedure:

(1) The mixing tool is correctly positioned

(2) The mixing shaft penetrates to the desired depth of treatment with
simultaneous disaggregation of the soil by the mixing tool
(3) After reaching the desired depth, the shaft is withdrawn and at the same
time the binder in granular or powder form is injected into the soil
(4) The mixing tool rotates in the horizontal plane and mixes the soil and
binder
(5) Completion of the treated column

10.2.2.1 Phase 1: Penetration of the mixing tool

In the irst phase the mixing tool penetrates the soil while it is rotated to
the required depth. The soil structure is remoulded during this penetration.
The magnitude of the remoulding depends on the penetration and rotation
440 Ground improvement

PF Pressure feeder
Excavator

Peat,

2-5 meters
Mass stabilished peat,
mud,
mud or soft clay
soft
clay

3-5 meters Preloading embankment Geotextile

(reinforcement)

Figure 10.4 Installation sequence for mass mixing with rotated cylinders and cutter
heads. (Courtesy of Allu Finland.)

speed of the mixing tool. Typical penetration speed is 100 mm/rev. and
typical rotation speed of the mixing tool is <100 rev/min. The binder is sel-
dom injected during penetration. In mass mixing using cutter heads there
is no initial penetration—instead the mixing and dispersion of binder take
place directly.

10.2.2.2 Phase 2: In-situ mixing and dispersion of binder

The formation of an in-situ mixed column takes place during withdrawal
of the mixing tool. The process starts by delivering binder via compressed
air from the powder tanks via hoses connected to the Kelly bar by a swivel.
The mixing tool, connected to the lower end of the Kelly bar, has an outlet
port or ports commonly at the top level of blades, see Figure 10.1, where
the binder is distributed into the soil. The binder is spread through the
cavities formed in the soil by rotation of the mixing tools upper level of
blades. The spread of binder through the column diameter depends on the
air pressure, the soil rheological properties, the geometry of the mixing
tool and the diameter, and the rate of withdrawal and rotation. In order
to increase mixing energy into the soil, tools with 2–4 levels of blade are
commonly used.
Before the uplifting starts and the creation of a column begins, the mix-
ing tool at the column base level is rotated just prior to lifting in order
to reach the required rotation speed and to inject the required amount
of binder in the toe of the column. This process takes approximately 10
seconds. Lifting of the mixing tool during high-speed rotation is then
commenced. Typical rotation speeds are 120–180 rpm and lift speeds are
15–30 mm/rev. This means that production speed varies between 1.2–5.4
 Dry soil mixing 441

m of column per minute. The installation sequence for DSM columns is

presented in Figure 10.3.
The distribution of binder through the column diameter signiicantly
depends on the disposition of the outlet port in combination with the
design of the upper level of blades on the mixing tool. Column installation
commonly stops 0.3–1.0 m below the working platform. This means that
the upper 1 m of a column should not be treated as a full-strength column.
In mass mixing using cutter heads, the distribution of binder takes place
from an outlet port in the bottom of the mixing unit, see Figure 10.2.
The cylinders (or drums) with the cutter head mounted rotate in a differ-
ent sense to aid mixing of the soil. Mixing is carried out in a site-speciic
pattern, commonly in cells of 3 × 3 m, until the required depth is reached.
Mixing energy in the soil, and distribution of binder evenly into the cell,
is highly dependent on the skill of the operator. Mass mixing can be per-
formed right to the ground surface. However, the top 0.5 m is often poorly
mixed compared with the mixed soil at greater depth. Installation sequence
for mass mixing is presented in Figure 10.4.

10.2.2.3 Phase 3: Molecular diffusion after installation

After the manufacture of a column or when mass mixing is completed, the
molecular diffusion takes place in the mixed soil volume. Molecular diffu-
sion of calcium ions from stabilised soil migrates into unstabilised surround-
ing soil, or from regions of stabilised soil with high concentration of calcium
ions into regions with poor concentration. The increase in strength caused
by the migration of calcium ions within a column has been poorly investi-
gated. However, Axelsson and Larsson (2003) reported that observations
on extracted lime-cement columns showed that the columns seem to heal
a short time after column penetration tests, which may possibly be due to
migration of calcium ions. Mitigation of calcium ions from column periph-
ery has been the subject of many investigations (Rogers et al., 2000a, b;
Hayashi et al., 2003; Rogers and Glendinning, 1996; Löfroth, 2005). These
investigations show that calcium ions migrate approximately 20–30 mm
within 1 year and 40–60 mm within 10 years.

10.2.3 Factors important in the mixing process

The mixing process is complex so the result and quality of an in-situ mixed
column or the stabilised soil mass depends on a number of factors, after
Larsson (2005). Signiicant issues include

r The rheological soil properties

r The type of binder and the dosage
r The mixing energy
442 Ground improvement

r The mixing tool design and the drilling rig site–speciic adjustments
r Air pressure delivery and amount of air entrained into the ground

10.2.3.1 Rheological soil properties

The rheological soil properties have a signiicant inluence on the eficiency
of the mixing and quality of a mixed soil volume. Cohesive soils with mod-
erate to high water content have, by their nature, a considerable resistance
to remoulding compared with cohesionless soil. In soft soils the natural
water content is often near the liquid limit, and the incorporation of dry
binder rapidly reacts with the soils’ natural water and becomes more plas-
tic and more resistant to remoulding. Incorporation of a dry powder will
signiicantly change the rheological soil properties by dewatering the soil
volume that is being mixed.

10.2.3.2 Type and amount of binder

The inluence of type and quantity of binder affects the mixing process.
The fact is that increasing the amount of binder will increase the strength
of the soil, with some exceptions. On the other hand, increased amount
of binder can decrease the production speed due to larger quantities of
material to be transported in the feeding system and larger quantities need-
ing dispersion into the soil. Very large amounts of binder can have a neg-
ative effect on the strength value due to inadequate water content. The
relationship between water content and binder content, water/cement ratio
(w/c), has been investigated in a number of studies (Åhnberg et al., 1995;
Babasaki et al., 1997; Filz, 2012) and is an important factor to evaluate in
the pre-design stage. The water/cement ratio and the relation to a certain
strength gain are soil speciic. Figure 10.5 shows the relation between shear
strength and w/c-ratio.
The w/c ratio in soil mixed samples can be expressed as:

 ρsoil 
 ρsoil − (1 + w ) 
w/c =  N 
(10.1)
mbinder
where
ρsoil = bulk density of unstabilised soil (t/m3)
wN = natural water content in unstabilised soil
mbinder = binder content (t/m3)

Graphs on the relationship between w/c ratio and strength from labora-
tory investigations and from literature studies are valuable information in
the pre-design stage. However, mixing soil samples with different amounts
 Dry soil mixing 443

Shear strength
kPa
300
Clayey mud
28 days
250

200
Cement

150
Lime cement

100

50

0
0 4 8 12 16
wcr-wctr

Figure 10.5 Relation between shear strength and water cement ratio (w/c) in a clayey mud
from Sweden. (From Åhnberg, H., Johansson, S.E., Retelius, A., Ljungkrantz,
C., Holmqvist, L., and Holm, G. (1995). Cement och Lalk för Djupstabilisering av
Jord – En Kemisk Fysikalisk Studie av Stabiliseringseffekter (Cement and Lime for
Stabilisation of Soil at Depth – a Chemical Physical Investigation of Soil Improvement
Effects), Report No. 48. Linköping, Sweden: Swedish Geotechnical Institute.)

of binder in laboratory is commonly performed in order to investigate suf-

icient amounts of binder. Increased binder content increases the strength
gain to a certain limit; thereafter, strength gain is decreasing due to lack
of water in the natural soil. On the other hand, if the binder content is
decreased, that would mean that at a certain level there is no or very limited
strength gain in the mixed soil (also known as a threshold condition).

10.2.3.3 Mixing energy

Mixing energy is a measure of the mixing work that the mixing tool
causes during the installation process. The degree of mixing work imple-
mented into a soil is related to the column strength and the dispersion of
the binder across the column area (Muro et al., 1987a, b; Nishida et al.,
1996; Larsson, 1999; Larsson et al. 2005a, c). Increased mixing work has
an increasing effect of the column strength in the soil. However, it is not
possible to predict column strength based on the strength of information of
a certain mixing work.
444 Ground improvement

Mixing energy is, in its more general sense, measured as mixing cycles
per metre of a column T (Yoshizawa et al., 1997):

 N d   N u  
T = ∑ M ×  V  +  (10.2)
 d   Vu  
where:
ΣM = number of mixing tool blades
Nd = rotation speed of mixing tool during penetration (rev/min)
Vd = mixing tool penetration velocity (m/min)
Nu = rotation speed of mixing tool during retrieval (rev/min)
Vu = mixing tool retrieval velocity (m/min).

For DSM, by the Nordic method, the uplift rate (mm/rev) of the mix-
ing tool is used as a measure of the mixing time since mixing does not
occur during penetration. The mixing energy is speciied as the blade
rotation number (BRN; see Figure 10.6) and the number of cycles per
column metre T is calculated as:

1
T = ∑ M × s × 1000 (10.3)
Strength magnitude and variability

Strength

Coefficient of variation

Blade rotation number, T: n/m

Figure 10.6 Principal changes in strength and coeficient of variation due to the variation
in blade rotation number T. (From Larsson S. (2005). Keynote lecture: state
of practice report – execution, monitoring and quality control. Proceedings of
the International Conference on Deep Mixing Best Practice and Recent Advances,
May 23–25, 2005, Stockholm, Sweden, pp. 732–785.)
 Dry soil mixing 445

where
M = number of mixing tool blades
s = retrieval rate of mixing tool during withdraw (mm/rev).

Increased mixing energy is favourable due to increased magnitude of

strength and decreased variability in the mixed soil. However, a high degree
of mixing energy reduces the production speed of a project. Negative effects
of the column quality can also be a consequence of a high degree of mixing.
For example, a high degree of mixing means that the mixing tool spends
more time in the ground resulting in more compressed air into the soil. This
can generate heaving and a phenomenon causing ‘craters’.
The magnitude of mixing energy needs to be evaluated for each speciic
site and the site-speciic type of soil. Laboratory mixing tests gives an indi-
cation of the required magnitude of mixing energy, which will be veriied
(or adjusted if necessary) during the installation of test columns before the
production begins.

10.2.3.4 Mixing tool design and drilling rig

site–specific adjustments
Adjustments of the drilling rig to suit the site-speciic soil properties are
important in DSM projects. This includes evaluation of mixing tools,
adjustments of rotation speed, evaluation of mixing energy (testing differ-
ent retrieval rates), and adjustments of air pressure in the feeding system.
These parameters and the adjustments are discussed in Section 10.3.

10.2.4 Binders and soil properties

10.2.4.1 Binders
Binders used for DSM are cement (standard Portland cement), lime (quick-
lime), slag (granulated blast furnace slag), ly ash, and gypsum. There have
also been some investigations of other binders, primarily in laboratory test-
ing; however, these additives are not used extensively. Indeed, the most
commonly used binders are cement and quicklime. These are often pre-
blended to gain the most suitable binder for the site-speciic soil properties.
The local access to different additives and regional industrial manufactur-
ers (access to cost-effective material) also inluences the choice and blend
for site-speciic binders. Notwithstanding general experience available,
laboratory mixing tests on soil samples from the site to be treated are an
essential component in the selection of binder and binder blends.
In the Nordic countries, pre-blended binders of quicklime and cement
are by far the most commonly used binders. Pre-blended mixtures of
lime/cement and ly ash are moderately used as are pre-blended mixtures
446 Ground improvement

of cement and slag. In other European countries, the UK, and the US,
the most commonly used binders are cement only or cement and slag.
However, blended mixtures of cement and quicklime are moderately used
in Poland.

10.2.4.2 Applicable soil types

Dry soil mixing is applicable in soft soils with high moisture content. As the
dry binders need in-situ water for the chemical reaction, a certain amount
of evaporable water is necessary. As a guideline, a minimum moisture con-
tent of 30% is necessary for DSM. Table 10.1 provides a summary of bind-
ers that are commonly used in different soil types.
The presented amounts of binders are for soft to medium soft columns
with a design undrained shear strength of 150 kPa or less.
The described binder in Table 10.1 is one, two, or three component
blends. In almost every blend cement is the basic additive and lime, slag, ly
ash and gypsum are secondary additives. Historically, in the Nordic coun-
tries lime is the basic binder and other admixtures have been developed
from there.

10.2.4.3 Binders and reaction process

The reaction process and the function of different binders have been investi-
gated in numerous reports. The investigation by Janz and Johansson (2002)
is recommended for a detailed description of the reaction process. In this
chapter the most common used binders—cement, lime, blast furnace slag
(GGBS) and ly ash—are described.

10.2.4.3.1 Cement
Cement used as binder in a DSM process is predominately standard
Portland cement. When cement is mixed with a soil it reacts with the water
content and instantly the hydration process commences in which a hard
cement paste forms of calcium silicate hydrate Ca3Si2H4 (CSH-gel). The
CSH-gel is formed on the cement particles and increases in size, illing the
voids between the particles. With time the porosity decreases, the particles
bind together, and the mass becomes stronger and more dense. Initially, the
rate of strength gain is controlled by the temperature; the higher the tem-
perature, the more reactions that take place leading to better strength gain
(Timoney et al., 2012). The reaction process of 100g cement and 25g H2O
can be expressed as:

Cement + H 2O → CSH-gel + Ca(OH)2

100 g 25 g → 100 g 25 g
 Dry soil mixing 447

Table 10.1 Soil types and binders

Soil type Binder Amount of binder Commentary
Clay Lime, Cement, Lime/ 70–110 kg/m3
Cement, Lime/ Pure cement
Cement/Fly Ash 90–150 kg/m3
Quick clay Lime, Lime/Cement, 70–100 kg/m3 Quick reaction,
Lime/Cement/ especially with high
Fly Ash lime content.
Silty clay Lime/Cement, 70–110 kg/m3 High degree of
Cement, Cement/ cementation with
Slag cement.
Organic clay Cement, Lime/ 100–200 kg/m3 Slow reaction, minor
Cement, Cement/ part with lime
Slag commonly speeds
up the reaction.
Sludge Cement, Cement/Slag 120–250 kg/m3 Slow reaction.
Dificult to predict
strength increase.
Clay with high Cement, Lime/ 120–250 kg/m3 Slow reaction. Large
sulphide content Cement, Cement/ variations in strength
Slag gain. Local knowledge
important.
Silt Cement, Cement/ 100–150 kg/m3
Slag, Lime/Cement
Sandy silt Cement, Cement/Slag 60–110 kg/m3 Natural moisture
content needs to be
larger than 30%.
Peat Cement, Slag/Cement 150–>300 kg/m3 Very important with
ield and laboratory
tests.
Dredged material Cement, Cement/ 70–110 kg/m3 Tests necessary,
(Mud) Slag, Cement/Fly Ash especially due to
contaminations.
Contaminated soils 70–110 kg/m3 Tests necessary,
especially due to
type of
contamination and
leakage tests.

10.2.4.3.2 Lime
Lime used as binder is mainly calcium oxide (CaO), known as quicklime or
burnet lime. When quicklime is mixed with water slaked lime or hydrated
lime, Ca(OH)2 , is formed:

CaO + H 2O → Ca(OH)2
100g 32g → 132g
448 Ground improvement

The reaction is instant, and reaches its maximum within 5 minutes while
generating a great deal of heat. The reaction also results in the pH increas-
ing to ~12.5, which is a condition for the secondary pozzolanic reaction.
During soil stabilisation, hydration dewaters the soil giving a rapid gain in
stability (Janz and Johansson, 2002). The primary reaction is not gaining
any strength. The strength gain in stabilised soil is due to the secondary
pozzolanic reactions with other additives or with the surrounding soils:

Ca(OH)2 + pozzolana + H 2O → CSH (CASH)

Humic acid inhibits strengthening reactions, which leads to poor strength

gain performance in organic soils.

10.2.4.3.3 Granulated blast furnace slag (GGBS)

Ground GGBS is a by-product of the steel manufacturing processes. GGBS
is a latent hydraulic material, which means it contains lime but requires
an activation before it can react with water. GGBS is used as a secondary
binder in soil stabilisation. The temperature created during the reaction is
low, which results in slow strength gains. In the Nordic countries GGBS
blended with cement has been used successfully in mass mixing projects
(Jelisic and Leppännen, 1999).

10.2.4.3.4 Fly ash

Pulverised ly ash (PFA), similar to GGBS, is also a latent hydraulic material
that either can be obtained from lue gas in the coal-ired power industry or
from the paper mill industry. The lue gas from the power generation plants
is a ine-grained material with pozzolanic qualities. The pozzolanic reaction
in PFA requires an activator as cement or lime to obtain strength gain. PFA,
lime, and cement in the proportions 33%/33%/33% is commonly used in
Sweden as a binder alternative to lime/cement 50%/50% in marine clays.

10.2.4.4 Strength gain

Strength gain and the magnitude of strength gain in stabilised soil depends
on several factors, such as soil characteristics, type of and amount of
binder, temperature, time aspects, and stress situation. The inluences of
these factors in the strength gain process differ between binder combina-
tions and soil types.

10.2.4.4.1 Soil characteristics

As shown in Table 10.1, different types of soil are more suitable for certain
types of binders or binder combinations. The required amount of binder to
 Dry soil mixing 449

gain a certain magnitude of strength depends predominately on the natural

moisture content and the organic content in the soil. Soils with high organic
content (>6%) require large amounts of binder (not necessarily higher w/c
ratio compared to inorganic soils).
In the mixing process, some general soil characteristics can be identiied;
soils with high sensitivity (for example quick clay) are easier to mix than
soils with low sensitivity; soils with high organic content (sludge, mud)
often have a slow increase in strength compared to inorganic clay.
Timoney et al. (2012) reviewed published data on a wide range of peats
and highly organic soils and found that the 28-day unconined compression
strength (UCS) showed some correlation with the ratio of water to binder
by weight in the mix. Signiicantly, the amount of humiication was also
highly relevant with a clear inverse relationship found between UCS and
humiication using the von Post* classiication.

10.2.4.4.2 Type of binder and amount of binder

As described, different binders react differently in the same type of soil.
An outline of the strength gain in clay and silty clay is shown in Figure
10.7. Typical amount of binders in different type of soils are presented in
Table 10.1. Local experience and results from laboratory mixing tests are
important parts in determining a suficient amount of binder. Site-speciic
requirements shall also be evaluated; for instance, the need for rapid
strength gain in order to use the installed column shortly after installation;
and the interaction with unstabilised soil especially when used to increase
the stability in natural slopes.

10.2.4.4.3 Temperature
Soil masses stabilised with lime and/or cement generate heat during the
cement and pozzolanic reactions; Halkola (1999) reports a temperature
of 70°C in lime and Åhnberg et al. (1995) reported that cement increases
the ground temperature by 5–10°C and lime can increase the ground
temperature 40–50°C, and locally it can even generate temperatures up
to 100°C. Binders like GGBS produce less heat during exothermic reac-
tions, and are therefore more susceptible to temperature changes of the
soil being stabilised, thus often resulting in less reaction and lower initial
strength (Timoney et al., 2012). Admixtures with lime or cement increase
the ground temperature, which activates any secondary binder such as
GGBS or ly ash.

* von Post is a classiication system for peat on a scale between H1 to H10 based on its physi-
cal properties.
450 Ground improvement

Increase of shear strength over time

700

600

500
Shear strength (kPa)

400

300
Lime

200 Lime/cement

100 Cement

0
0 50 100 150 200 250 300
Time (days)

Figure 10.7 Strength gain over time for different binders. (From Åhnberg, H.,
Johansson, S.E., Retelius, A., Ljungkrantz, C., Holmqvist, L., and Holm,
G. (1995). Cement och Lalk för Djupstabilisering av Jord – En Kemisk
Fysikalisk Studie av Stabiliseringseffekter (Cement and Lime for Stabilisation
of Soil at Depth – a Chemical Physical Investigation of Soil Improvement
Effects), Report No. 48. Linköping, Sweden: Swedish Geotechnical
Institute.)

10.2.4.4.4 Pre-stress loading

Strength gain in peat is dependent on the pre-stress level during curing.
Laboratory tests carried out by Åhnberg (2001) investigated the effect on
pre-stress loading on stabilised peat samples. Samples were loaded with 0, 9,
and 18 kPa at 45 minutes, 4 hours, and 24 hours after mixing, respectively.
It was observed that pre-stress loading within 45 minutes increased the
strength in the samples up to several times compared to unloaded samples.
Compression that occurs during preloading reduces the distance between
the binder grains and the particles in the peat and facilitates building
bonds in the stabilised soil mass. Hebib and Farrell (2003) showed from
tests on Irish peat that the permeability of the stabilised samples were also
reduced by pre-stressing, whereas the permeability of samples not subject
to pre-stress were the same as for unstabilised peat.
In the ield, strength gain of stabilised peat (especially mass mixed) needs
to be pre-stressed by approximately 1 m of ill in order to compress the
loosened soil volume, which has been remoulded by the mixing tool and the
compressed air. Pre-stress load is recommended to be placed on the mixed
soil as soon as possible after mixing; however, the common requirement is
that the load shall be applied within 24 hours after mixing. Pre-stressed
 Dry soil mixing 451

loading on DSM columns in clay has also have a positive effect on the
strength gain, but less than in peat.

10.2.4.4.5 Prediction of strength gain

The time dependency of strength gaining and prediction of strength increase
according to type of binder and soil conditions has been investigated in a
number of reports, such as Nagaraj et al. (1996), Porbaha et al. (2000),
Horpibulsuk et al. (2003), and Åhnberg (2006). For cement-stabilised soils,
Åhnberg (2006) investigated strength increase in clay and sludge samples
tested between 7 and 800 days, see Figure 10.8.
Prediction of strength gain for cement-stabilised clay as:
qt
= 0.3 × ln t (10.4)
q28
where
qt = UCS after t days
q28 = UCS at 28 days
t = time (days).

3
Linköping clay
y = 0.305ln(x) – 0.04
Löftobro clay
R2 = 0.94
Holma gyttja
Trend Nagaraj et al. (1996)

2
qt/q28

Horpibulsuk et al. (2003)

0
0 200 400 600 800 1000
Time after mixing t, days

Figure 10.8 Increase in unconined compression strength over time for cement-stabilised
clay and sludge. (From Åhnberg, H. (2006). Strength of Stabilized Soils – A
Laboratory Study on Clays and Organic Soils Stabilised with Different Types of
Binder. PhD Thesis, Lund University, Sweden.)
452 Ground improvement

Prediction on strength gain for cement-stabilised soil has also been investi-
gated by Horpibulsuk et al. (2003) by using a modiication of Abram’s law*.
Predictions of strength gain by time for cement-stabilised clays are dif-
icult and the presented methods are empirical methods that can be used
as an indication of the strength gain over time. However, these can only
provide indications while site-speciic tests are the most reliable method.

10.2.4.5 Long-term strength

Long-term strength and durability have been investigated by ield tests and
laboratory tests. For Japanese experience, Terashi (2003) summarised the
results from a number of studies. The results demonstrate that signiicant
increases of strength have occurred from the 28 days’ strength to 10-year-
old lime and cement-treated soils. Minor leaching of calcium (CaO) ions
that can decrease strength has been observed at the periphery of the col-
umns. However, the leaching was slow and small in scale, and the author’s
conclusion was that the leaching is much less than the strength gained over
time. Löfroth (2005) reported similar results on 9.5- to 11-year-old lime/
cement columns installed in organic clay and highly sensitive clays on the
Swedish west coast.
Strength gain continues over the long-term due to migration of calcium
ions and pozzolanic effects in the columns. The above investigations showed
an increasing strength 10 years after installation. However, the magnitude
of the increase depends on the site conditions as well the binder type.

10.3 EQUIPMENT, MONITORING, AND CONTROL

The development of new equipment and the modiication of existing equip-

ment is ongoing. Description of the execution procedure of the Nordic method
was recently presented by Bredenberg (1999) and Larsson (2003, 2005).

10.3.1 Equipment
One machine unit consists of a drilling rig and a shuttle carrying the binder
tank (some machines have the binder tank on the drilling rig). The drilling
rig is constructed with wide tracks for low ground pressure (40–60 kPa).
Such rigs are 4–5 m long and 3.0–3.5 m wide and are commonly the
base units of excavators or a piling rig, see Figure 10.9. The mast is nor-
mally 15–17 m high, but can be extended up to about 27 m. The machine
equipment and shuttles are commonly manufactured by the contractors
themselves.

* Abram’s law is an empirical model to predict strength gain extensively used in concrete
technology.
 Dry soil mixing 453

Figure 10.9 DSM machine unit, drilling rig, and shuttle. (Courtesy of dmixab.)

The shuttles carrying binders have a storage capacity of 10–15 tons and
are pressurised to 5–10 bar. The pressures applied to the ground by the
shuttle are similar to those from drilling rigs. The binder is fed from the
shuttle to the top of the drilling shaft and blown out by the outlet hole at
the mixing tool via air pressure, see Figure 10.1. Drilling can generally, with
equipment today, be carried out to approximately 25 m depth; however, in
the majority of projects drilling length is 15 m or less. Rotation speed of the
drilling shaft is in the interval of 100–200 rpm, depending on the ground
conditions. Downward penetration is usually controlled manually by the
operator and corrected due to ground conditions and the risk of hitting
obstacles. Typical downward movement is 100 mm/rotation. During down-
ward movement, the shaft is pressurised by air to prevent water and soil
from entering. During upward movement, the binder is fed out by air pres-
sure. The amount of binder is regulated by feed-out valves in the shuttle.
The upward movement is generally 15–30 mm/rotation, depending on the
required mixing work. The blowout of binder into the surrounding soil is
stopped at 0.5 m beneath the ground surface to prevent release into the
open air. In urban areas a protection head can be used to avoid clay splash
and blow outs of binder into open air.
Mass mixing equipment is similar to DSM equipment with a drilling unit
(rig and shuttle). The drilling rigs are commonly smaller than a DSM rig
and have lower ground pressure (30–40 kPa), see Figure 10.10. Shuttles are
in many cases not used for binder delivery due to limited storage capacity
454 Ground improvement

(a) (b)

Figure 10.10 (a) Mass mixing drilling rig and shuttle. (Courtesy of LCM.) (b) Storage sup-
ply in bulk silos. (Courtesy of LCM.)

(large amount of binder is commonly the case in mass mixing). The binder
is delivered directly from bulk silos by compressed air in tubes or supplied
by bulk trucks behind the drilling rig.
The mixing tool is an important part of the equipment. There are some
well-established shapes of mixing tools. The three most commonly used
are (1) standard mixing tool (the irst developed mixing tool, Figure 10.11);

Figure 10.11 Standard mixing tool, Pinnborr (shown in Figure 10.1), and Allu rotary head
(shown in Figure 10.2).
 Dry soil mixing 455

(2) Pinnborr, a three-level blade mixing tool (the most common in Nordic
countries today); and (3) Allu rotary head (mass mixing tool).
Since the turn of the century, industrial mixing has made progress as a
new technique based on knowledge gained from DSM experience. Industrial
mixing is predominately used for dredge material pumped into the machine
where it is mixed with different types of binder and thereafter transported
to a land reclamation area (a lagoon) for hardening.

10.3.2 Monitoring and control

Monitoring and control of the installation process are very important since
the stabilisation process itself seldom leads to direct inspection. The amount
of binder injected as well as the geometry and homogeneity of the stabilised
soil volume, whether it is columns or mass-mixed soil, must be evaluated by
indirect measurements of binder use, retrieval rate, rotation speed, and so
on. The technical problems related to monitoring are more pronounced for
the dry than for the wet method since the binder is distributed by a com-
pressed air stream in the dry method. Therefore, the weight loss of binder in
the storage tank is used as a measure of the injected amount of binder. Load
transducers continuously monitor (20 readings per second) the weight in the
storage tank and transfer the information to the machine computer.
Each contractor has their own monitoring system for control, recording,
inspection, and documentation of the improvement work, but these param-
eters at a minimum should be recorded for every DSM project:

r Column name
r Machine name
r Responsible operator
r Used mixing tool
r Time and date of installation
r Drilling time
r Retrieval rate (mm/r)
r Speed of rotation (rpm)
r Drilling length and/or stabilising length
r Amount of binder along the column
r Weight of the binder tank
r Position of the column (GPS)

Manual registration by the machine operator parallel to the computer reg-

istration is recommended. The following is registered on manual logs:

r Machine name
r Responsible operator
r Date of installation
r Column name
456 Ground improvement

r Drilling length and/or stabilising length

r Deviations and other obstacle
r Service, change of mixing tool, and so on

Monitoring data is part of the quality assurance (QA) and quality control
(QC) for the stabilisation work. The information is commonly delivered
from the contractor to the client on a daily or weekly basis. Adjustments to
the installation process and decisions for in-situ testing can be made from
the printouts of the installation data. It is important not to focus solely on
performing in-situ tests on apparently less good columns, as it is the mass
properties that will often be most relevant.
The production capacity is dependent on a number of factors and site-speciic
conditions such as stiffness in the unstabilised soil, magnitude of obstacles,
length of columns, rotation speed and restrictions in vertical movement (special
installation pattern etc.), access to the area and transportations roads, distance
to binder storage, and amount of binder in the columns. Typical production
capacity for 10-m long columns is 40–80 stabilised metres per hour per unit.

10.4 APPLICATION, DESIGN AND TESTING

10.4.1 Applications
There are a number of applications using DSM columns and mass mixing
for permanent and temporary works either on land or in marine environ-
ment. Some of the main applications include

r Reduction of settlement
r Improvement of stability
r Reduction of ground vibration
r Foundations of structures and houses
r Mass mixing of organic soils and dredged sediments
r Solidiication and stabilisation of contaminated soils

Secondary applications include the following:

r Increase passive earth pressure for sheet pile walls in soft clay
r Reduce active earth pressure on retaining/sheet-piled walls in soft clay
r Preventing liquefaction in seismic hazard areas
r Creating geohydrological barriers
DSM columns and mass mixing is sometimes combined with other soil
improvement techniques in order to design the most technical and eco-
nomical solution:

r DSM columns combined with light ill aggregates or expanded poly-

styrene (EPS) material in the embankment. This combination is
 Dry soil mixing 457

commonly used for embankments in the transit zones of bridges and

piled structures.
r DSM columns and wick drains are occasionally used, especially in
areas with deep deposits of clay. In these applications DSM columns
are mainly installed to improve the stability of the embankment and
wick drains are installed to reduce settlements at depth.
r DSM columns and vibro replacement have also been combined,
(Dahlström, 2012). DSM columns increase the coninement and shear
strength in the soil. The improved ground thereby increases the load
capacity of the vibro replacement columns.

Depending on the application, column spacing (or improvement ratio), lay-

out, diameter of column, and length of column all require careful attention.

10.4.1.1 Reduction of settlements

Reduction of settlements for road and railroad embankments, parking
areas, and areas around structures are the most common application of
DSM columns and mass mixing. In soft soils with shear strength greater
than 8 kPa, columns are installed as single elements with a spacing of 1.3–
3.0 times the column diameter. In extremely soft soils (shear strength less
than 8 kPa) the coninement of the columns are limited, hence the col-
umns need to be installed in panels or grids to support each other. In these
extremely soft soils mass mixing is an alternative to interlocking columns
in panels or grids. A combination of DSM columns and mass mixing is
commonly used in areas with extreme soft soil in the top 1–5 m followed
by soft clay deposits (e.g, Dahlström and Eriksson, 2005).
Columns are mainly installed into a irm soil layers in order to distribute the
load from the embankment. In very deep soft soil deposits, and a low to mod-
erate height of the embankment, columns are installed to a predesigned depth.
The improvement ratio can also vary by depth (e.g., every second column is
installed to greater depth) in order to provide the most economical solution.
Arching between single elements such as columns needs to be checked,
especially in the case of low embankments and column spacing greater than
twice the column diameter. Geogrids or load transfer platforms can be used
to secure arching between the elements.
The magnitude of reduction of settlements mainly depends on the
improvement ratio and column strength. Generally, settlements are reduced
2–5 times compared with unimproved soil.

10.4.1.2 Improving stability

Columns can be connected with each other to create a panel or a grid of
interlocking columns (see Figure 10.12) for improving the stability of road
458 Ground improvement

Figure 10.12 Interlocking columns in panels at Bärbyleden, Uppsala, Sweden.

and railroad embankments, slopes, and temporary excavations. The spac-

ing between columns in a panel is generally in the range of 0.75–0.85 times
column diameter. The connection between columns is the most critical part
of a panel, and shearing between columns is the most common cause of
failure. Panels are installed perpendicularly to the most critical failure plane
and have a length along the top to ensure that the total area of improved soil
does not slide as a mass block. Panel length must be enough to mobilise suf-
icient shear resistance in the panel. In slopes, panels installed with inclina-
tions of 10:1 to 5:1 are advantageous due to the increased axial load on the
panels. Panel depth depends on the most critical failure plane in stabilised
and unstabilised conditions, and both cases therefore need to be investi-
gated. Overlapping of columns with great depth is extremely dificult to
perform with the equipment available at the time of writing. A recommen-
dation by the Swedish Road Authorization is that overlapping of columns
deeper than 8 m shall, without investigations, be used only with restrictions.
The distance between panels depends mainly on three parameters: the
width of the panel, the stability of the unstabilised soil between panels
(squeezing), and the interaction between stabilised and unstabilised soil.
Typical spacing between panels is 1.0–3.5 times column diameter. Panels
can be installed as single panels, which is the most common design, or as
two interlocking panels. Interlocking panels are sometimes used in areas
where the function of the panel is very critical, especially at depth.
Construction of columns that will form a panel requires higher require-
ments of precision and inclination during installation. Tests using vertical
 Dry soil mixing 459

inclinometers to control the declination after installation of columns in a

panel have been undertaken in some projects. Installation of panels in area
with low stability or for live railway embankments (Pye et al., 2012a, b)
needs to consider the effects of increasing pore pressure and movements
during installation.

10.4.1.3 Reduction of vibrations

Reduction of vibrations is predominately associated with high-speed trains
(>180 km/h) travelling over soft clay deposits. The train-induced ground
vibrations generate different waves (P-, S-, and R-waves) with different veloc-
ity propagating in the soil. When the train speed exceeds any of these wave
velocities, the character of the propagation of the waves is dramatically
changed. This phenomenon is called a shock front and will give rise to high
levels of vibrations in the soil with large displacements as a secondary effect.
Since the late 1990s, DSM columns have been used as soil improvement
to reduce the ground vibrations and displacements connected to the high-
speed phenomenon of a train passing over soft clay on low embankment.
Columns are installed in a speciic pattern, see Figure 10.13, which was
irst developed within a Swedish research project (Holm et al., 2002). In
their report, the measurements of displacements on the trackbed showed a
reduction of approximately 5 times at low speed and of approximately 15
times at high speed (200 km/h) compared with measurements taken before
installation of DSM columns.
The speciic layout was adopted in order to develop stiff ground below the
ballasted track structure, and also to create a barrier for the surface waves
travelling along the track. As shown in Figure 10.13, the longitudinal panels
are centred underneath the rails in order to create a stiffened base. These
columns are 6–8 m long (commonly 7 m) and are the most important part of
the structure. Therefore, these columns are installed irst in order to secure
a suficient interlocking. Thereafter the perpendicular panels connecting the
longitudinal panels are installed to minor depth, commonly 4–6 m. Finally,
the single columns for reducing settlements are installed. The critical part in
this system is the contact between column and track ballast. Hence columns
are exposed and visually examined before back illing of track ballast.
This application is today a standard method, adopted in Scandinavian
countries for construction of new high-speed railways on soft soils. Barriers
of DSM columns installed in 2–4 interlocking panels for reduction of
ground vibrations have also been used as protection of built-up areas.

10.4.1.4 Foundations of structures and houses

Buildings, warehouses, and smaller bridges are structures for which DSM
columns are used to improve the sub soil layers. Concerning smaller bridges
460 Ground improvement

Exposure of col.

Section

Panels 6–8m
longitudinal

Panels 4–6m
perpendicular

Single col.
settlements
Plan

Figure 10.13 Column layout for reduction of vibration connected to high-speed train
phenome.

and culverts, the adjoining embankments are also usually stabilised with
DSM columns. By improving the soil underneath the structure, a system
with small differential settlements is obtained. An example of column lay-
out is shown in Figure 10.14. The ratio of improvement is increased under-
neath the structure and at the adjacent embankment, but further from the
embankment the ratio of improvement decreases. Columns in panels are
installed to increase the stability toward the passway. Panels along the road
are installed to increase the stability during excavation and foundation
works. This type of layout takes care of the permanent situation as well as
the temporary excavation and foundation works.
Buildings, warehouses, and residential homes built in soft clay and silt
areas have been supported by DSM columns as an economical alternative
to piling. When it comes to warehouses, a combination of DSM and piling
has been used in areas with soft clay deposits. The warehouse framework
structure is supported on piles taken to a irm bearing soil layer, and the
loor is supported by DSM columns.
Designing DSM for supporting structures means that small settlements
and restrictions of differential settlements are critical. Hence the static load
 Dry soil mixing 461

(a)

Bridge

(b)

Figure 10.14 (a) Column layout applied on small bridges and culverts. (b) Interlocking
panels for slope stability at Skepplanda, Sweden (Courtesy of Johnny
Wallgren.)

(permanent load) on a column shall be limited to 0.4–0.6 times the bearing

capacity of a column.

10.4.1.5 Mass mixing and stabilisation of highly

organic soils and dredged muds
Mass mixing is used in extremely soft soils or soils with high compress-
ibility and organic content as peat, mud and in mangrove (see case history
for Jewish Creek). The majority of mass mixing projects are for roads and
462 Ground improvement

parking areas and low embankments (0.5–3 m) to a depth of 1 to 5 m. The

design, execution and working scheme of a mass mixed volume, in combi-
nation with DSM columns, is more or less the same for stabilising dredged
mud for land reclamation as it is for stabilising peat for a road embank-
ment, see Figure 10.15. Execution of mass mixing remoulds a large volume
of soil and mixes in a considerable amount of air from the process into the
mixed soil volume. To obtain the target strength, it is necessary to allow
the air to dissipate into the atmosphere, which is best achieved by starting
preloading within 24 hours. Installation is executed in square or rectangu-
lar cells of 15–25 m 2 at a time.
The performance will be carried out in the following steps:

r Working bed for the irst front

r Installation of LC columns (due to the construction)
r Mass stabilisation
r Placing a geotextile and/or a geogrid on the stabilised soil
r Filling 0.3–0.5 m of cross-material
r Placing a geogrid
r Filling 0.3–0.5 m of cross-material
r After approximately 1 month (28 days curing time), the ill to inal
level
r The surcharge 3–6 months before the inalising the embankment

Dahlström and Eriksson (2005) have reported two projects using mass
stabilisation and DSM columns for road embankments in Sweden. Jelisic

Road level
0.5-1m

Working bed level

Geogrid
Geotextile
1-5m

MS-soil

DSM

Figure 10.15 Mass mixing combined with DSM for a road embankment.
 Dry soil mixing 463

and Leppännen (2005) and Forsman et al. (2008) have reported experi-
ences of mass stabilisation in contaminated dredge mud.

10.4.1.6 Solidification and stabilisation of contaminated soil

Solidiication and stabilisation of contaminated soils (the so-called S/S
method) has been an increasingly common application in the last decade.
The S/S method is predominately executed by mass-mixing equipment in
limited cells or barge. DSM columns have also been used as barriers to
enclose the contaminated soil. Industrial mixing is a new up-and-coming
technique based on the knowledge from the DSM experience. The dredged
material is usually pumped into the machine where it is mixed with differ-
ent types of binder, and thereafter transported to the land reclamation area
(a lagoon) for hardening. This application is under development and stems
from the experiences with DSM.

10.4.2 Design
The design of the Nordic method and the Japanese method differ in their
basic philosophy. This chapter only presents the Nordic method as it has
been developed alongside the DSM processes.The original design was
developed for lime columns and was irst presented by Boman and Broms
in 1975 at the Nordic Geotechnical Conference, NGM-75, in Copenhagen.
Only minor changes and complementary theories have been developed since
the irst paper regarding the design method was published. In 2005 Álen
et al. (2005a) presented a new design method to determine the compres-
sion strength in the columns. Today, both the original design philosophy
developed by Boman and Broms and the new design model for determining
the compression strength and bearing capacity of a column is used in the
Nordic countries. The original design philosophy is still the most com-
mon design method. Design guidelines such as EuroSoilStab (2002) and
SGF (2000) are based on Boman and Broms’ design philosophy. Stability
failure has been investigated and presented in numerous papers, including
Kitazume et al. (1996), Kivilö and Broms (1999), Terashi (2005), and Filz
et al. (2012).
DSM columns and mass mixing are inhomogeneous to varying degrees,
with an irregular structure and varying properties. The columns and mass
mixing are primarily intended to interact with unstabilised soil at axial
loading. For other load situations such as horizontal loading (direct shear-
ing) or uplifting (tensile stresses), shear strength can be signiicantly lower
than measured. Columns subjected to tensile stresses shall be avoided.
DSM columns and mass mixing structures are designed for ultimate limit
state (ULS) and serviceability limit state (SLS).
464 Ground improvement

10.4.2.1 Design models

The design models are based on the assumption that the DSM column inter-
acts with the unstabilised soil, which implies that the design is based on
semi-hard columns. The assumption of interaction between columns and
unstabilised soil is based on the assumption that deformation in a column
is equal to deformation in unstabilised soil. The characteristic properties
of the stabilised soil volume can therefore be calculated using the ratio of
improvement. The ratio of improvement is deined as the area of a column
divided by the spacing between the columns (total area), see Figure 10.16.
The ratio of improvement for single elements can be expressed as:

Acol
a" (10.5)
Atot
where Acol = area of the column; Atot = total area, which can be calculated
from the effective diameter (De) according to column layout and spacing
between columns (s); Atot = πDe2 /4.
Soil characteristics such as the shear strength, settlement modulus, and
permeability of the improved soil volume are calculated based on the ratio

0.866s
s

s s

(a) Rectangular spacing (b) Triangular spacing

De = 1.13s rectangular spacing

De
De = 1.05s triangular spacing

(c) Effective diameter

Figure 10.16 Ratio of improvement.

 Dry soil mixing 465

Undrained, drained and combined shear strength in stabilized soil

100
Drained shear strength
90 in DSM and soil
80
Drained shear strength
in DSM and undrained soil
Shear strength (kPa)

70
60
Undrained shear
50 strength in DSM and soil
40
30
Design shear
20 strength
10
0
0 25 50 75 100 125 150 175 200
Effective stress (kPa)
τfu τfd τfud τf,dimA τf,dimD

Figure 10.17 Critical shear strength due to in-situ stress situation in the stabilised soil
volume; the bold line is the calculated shear strength used in the stability analysis.

of improvement. The improved soil volume can be divided into three zones,
A, B and C, see Figure 10.17 (Álen et al., 2005a), which are:

A. Upper zone of 0.5–2 m, transition zone. In the top part of the col-
umns, the homogeneity and strength can vary considerably. Hence,
the column shall not be treated as a full-strength column. Spacing
between columns due to arching in the embankment ill needs to be
evaluated. Column spaces greater than 3 times the column diameter
should be used with caution.
B. Stabilised volume. Column and soil are assumed to have full
interaction.
C. Unimproved soil underneath the stabilised volume. Soft soil under-
neath the columns is considered according to basic soil mechanics.
For time-settlement analysis the drainage length has to be adjusted
due to the fact that DSM columns have higher permeability than
unimproved soil.

10.4.2.1.1 Column layout

The layout and the application of the DSM columns have signiicant inlu-
ence on the design and design calculations (Table 10.2). Columns and mass
mixing underneath embankments are predominately axially loaded. In
466 Ground improvement

these situations full interaction between column and soil can be assumed.
When columns are installed in natural slopes, excavations, or embank-
ments with a very low factor of safety in unstabilised conditions where
full interaction between column and soil is uncertain, columns need to be
designed and checked in undrained, drained, and combined shear strength.

10.4.2.1.2 Design ultimate limit state (ULS)

Design in ULS is recommended for characteristic values. However, accord-
ing to new design guidelines adopted in Europe (Eurocode), design with
partial safety factors is recommended. In this section, designs with charac-
teristic values are presented. The characteristic shear strength of the stabi-
lised soil volume is calculated as:

τ f = τ f ,col + (1 − a) × τ f ,soil (10.6)

where
a = ratio of improvement, see Equation 10.5
τf,col = shear strength in a column
τf,soil = shear strength in unstabilised soil.

Characteristic shear strength in undrained conditions is calculated as:

τ fu = Cu ,col + (1 − a) × Cu ,soil (10.7)

where
C u,col = UCS/2, (unconined compression strength).

Undrained shear strength greater than 150 kPa is not recommended for use
in the design guidelines.

Table 10.2 Recommendations of column layout for certain applications

Recommended layout
Design/Application Single columns Panels Grid Block/Mass mix
Embankments with FOS ≥ 1.0 X (x) (x) (x)
Columns in shear and passive zone - X X X
Natural slopes - X (x) (x)
Excavations, temporary works - X X (x)
Railways, high-speed trains - (x) X X
FOS, factor of safety for the unstabilised soil condition; X, primary use for the application; (x)
moderate use for the application; -, not recommended for the application.
 Dry soil mixing 467

Characteristic shear strength in drained conditions is calculated as:

τ fd = τ fd ,col + (1 − a) × τ fd ,soil (10.8)

where
τ fd ,col = c ´col + σ ´× tan(φ ´col ) (10.9)

c ´col = β × Cu ,col (10.10)

β = 0 to 0.5
φ´col = 30–35 deg.

τ fd ,soil = c ´soil + σ ´× tan(φ ´soil ) (10.11)

Characteristic shear strength in combined conditions is calculated as:

τ f ,comb = τ fd ,col + (1 − a) × Cu ,soil (10.12)

Combined analysis means that the drained shear strength in the columns
and undrained shear strength in unstabilised soil are used in calculations of
shear strength in the stabilised soil volume.
Combined shear strength is often the most critical factor in the stability
analysis for slopes and excavations, due to limited axial load on the col-
umns. In stability analysis the most critical shear strength in the stabilised
soil is due to the stress situation on the columns and the shape of the slip
surface (evaluation of columns in active, direct and passive shear zone).
Figure 10.17 shows an evaluation of the critical shear strength due to the
in-situ stress for the active shear zone and direct shear zone.

10.4.2.1.3 Seismic design and dynamic loads

Design for dynamic loads, such as with high-speed trains and mitigation of
earthquake ground deformations, has been performed successfully. Studies
from both Japan and Scandinavia shows that cyclic loads with a shear
stress level less than 0.4–0.6 times the undrained shear strength (measured
from laboratory triaxial and UCS tests) had no reduction of the undrained
shear strength in the samples. Japanese studies showed increased undrained
shear strength, Bengtsson and Karlsson (2006).
Triaxial cyclic load tests on laboratory mix samples are valuable infor-
mation in the design procedure for dynamic loads. Example of design for
mitigation of earthquake ground deformations is presented by Martin et al.
(1999) and an example of a design and ield study of the mitigation of
track and ground vibrations by high-speed trains is presented by Holm
et al. (2002).
468 Ground improvement

10.4.2.1.4 Design serviceability limit state (SLS)

Design in SLS involves models for settlement calculations, bearing capac-
ity, and dynamic loading required in the reduction of vibrations and
deformations arising from dynamic loads. The ultimate bearing capacity
of a column was originally proposed by Broms (1984) as a function of the
undrained shear strength in a column and the effective horizontal stress.
The guideline EuroSoilStab and SGF’s Report 2:2000 is based on these
functions:

σ f ,col = qu ,col + KP (σ v 0 + msoil × σ v − ucol ) + ucol (10.13)

where:
q u,col = UCS (unconined compression strength) ~2*C u,col
1 + sin φ '´col
K P = Coeficient of passive earth pressure, K P =
σ = Initial total overburden pressure 1 − sin φ '´col
v0
msoil = Factor of stress increase into the unstabilised soil due to applied
weight from unit loads (e.g., embankment). The ratio of stress
increase msoil = 0 to 0.5. The stress increase depends on the load
distribution on the unstabilised soil
∆σv = Applied unit load on the columns (e.g., embankment)
ucol = Pore water pressure.

The ultimate bearing capacity increases with depths and the UCS in a
column.
Creep strength is a commonly used term in design of DSM columns. The
creep strength is a function of the expected homogeneity of the column
depending on the mixing work, the virgin soil properties and type of binder.
For instance, organic soils and peats are more dificult to mix to achieve
high homogeneity in the column. Creep strength of a column is estimated at
65%–90% of the ultimate bearing capacity.
The creep strength can be calculated as:

σ creep = mcreep × σ f ,col (10.14)

where:
mcreep = 0.8–0.9 (columns in clayey silt, silty clay and sandy clay)
mcreep = 0.7–0.8 (columns in clay)
mcreep = 0.65–0.7 (columns in organic clay, peat and contaminated
soils).

If a small settlement is required, then a safety factor on the creep strength

should be applied. The safety factor is 0.7–0.8 times the creep strength
(σcreep).
 Dry soil mixing 469

10.4.2.1.5 Settlement calculations

The settlement of a column group is governed by the weighted average modu-
lus of elasticity of the column and the compression modulus of the unstabilised
soil.
h × σv
sgroup = ∑a × E + (1 − a) × Msoil
(10.15)
col

where:
Ecol = 50–300*C u,col
M soil = compression modulus of the unstabilised soil and depends on
the stress applied on the soil and the preconsolidation pressure
in the unstabilised soil between the columns.

In mass mixing settlement calculations are governed by the modulus of the

mass mixed soil. Due to the requirement of preloading (in order to gain
strength in the mixed soil volume) considerable settlement, up to 30%–
35% of the stabilised volume, can occur during curing.

h × σv
sMS = ∑ MMSl
(10.16)

where:
M MSl = 50–100*C u,MS

Young’s modulus in a column Ecol depends on the soil properties, shear

strength, and the stress level. For example, columns in organic soils shall
assume low values, and columns in inorganic clay and silty sandy soils can
be assumed to have high values. For mass mixing, compression modulus
M MS is in the range of 50–100*C u,MS .
Due to large variations of Young’s modulus in the columns and the dif-
ferences and uncertainties between samples prepared in the laboratory and
in-situ strength in a column, settlement calculations are recommended for
probable maximum and minimum values of Young’s modulus in a column.
Hence, monitoring of settlements and consolidation are of great importance
in a DSM project.
New models for calculation of settlement have been developed by Álen
et al. (2005a) based on ield studies on trial embankments on the Swedish
west coast and inite element calculations. Álen et al. (2005a) commented,
‘The material behavior of DSM columns can be regarded as similar to
a highly over consolidated clay or maybe a very low strength concrete.
Both descriptions highlight that it is the drained strength properties that
governs the behavior of the DSM column.’ The compression strength
(bearing capacity) can therefore be expressed with Mohr-Coulomb failure
criteria as:
470 Ground improvement

2 × cos(φ ´) 1 + sin(φ ´)
σ ´col = × c ´+ × σ ´h − σ ´v0 (10.17)
1 − sin(φ ´) 1 − sin(φ ´)

The horizontal stress situation in the soil can be expressed as:

σ ´ h = σ ´ h 0 + σ ´ h ,soil = σ ´ h 0 + 0, 5 × σ ´ v ,soil (10.18)

where:
σ'col = maximum increase of stress in a column
σ'h,0 = horizontal effective stress in-situ conditions
∆σ'h,soil = increased horizontal stress in unstabilised soil
∆σ'v,soil = increased vertical stress in unstabilised soil.

Stress distribution from the columns to the unstabilised soil has been
reported by Álen et al. (2005a). The investigation showed that the load distri-
bution from trial embankments on loating columns (columns not installed
to irm ground) was similar to load distributions in unstabilised soil. Álen et
al. (2005a) presented a modiied Boussinesq’s stress distribution with depth.
The authors also presented a load split model (see Figure 10.18) where part
of the total applied load is divided into two loads, q = q1 + q2. Load (q2) is
applied at top of the columns and load (q1) is transferred to the toe of the
columns. Stress distribution by depth is thereafter calculated for the two
cases. In applications with loating columns and large deposits of soft clay
(commonly the case in Sweden) this model has been successfully used.

10.4.2.1.6 Consolidation rate

When the effective stress in the soil is less than the preconsolidation pres-
sure, settlement will develop rapidly. When the effective stress in the soil
exceeds the preconsolidation pressure, then the rate of consolidation settle-
ment in the stabilised soil volume is calculated similar to vertically drained
soil. The permeability of the stabilised soil is 200–600 times the perme-
ability of the soil (EuroSoilStab, 2002). The rate of consolidation can be
calculated by Equation 10.17, after Barron (1948) and Hansbo (1979) and
modiied by Åhnberg et al. (1986):

 −2 × c × t 
U = 1 − exp  2 vh  (10.19)
 R × f(n) 
where:
U = degree of consolidation
cvh = coeficient of consolidation in unstabilised soil with respect to lat-
eral drainage, normally assumed to be equal to 2*cvv
 Dry soil mixing 471

q = q1 + q2

q2

A
1:X

1:X
B

q1
1:2

C 1:2

Figure 10.18 Conceptual zones, principle of load split model and stress distribution from
a stabilised soil volume.

cvv = coeficient of consolidation in unstabilised soil with respect to verti-

cal low
t = period of consolidation
R = radius of inluence.

When columns are installed at distance c between the centre in a square grid,
the inluence radius can be expressed as R = 0.56*c. If the columns are installed
in a triangular grid, the inluence radius can be expressed as R = 0.53*c.

n2  1  1    n2 − 1 1 k 
f (n) = ×  ln(n) − 0 .75 + ×  1 − 2
+  2 × 2 × soil × L2D 
n −1 
2
n 2
 4n    n r kcol 
(10.20)

where:
n = R/r
r = column radius
472 Ground improvement

c = distance between column centres

L D = column length with drainage upwards only, and half column
length with drainage both upward and downward
ksoil = permeability of unstabilised soil
kcol = permeability of column.

Investigations in Finland and Japan show that the permeability in the col-
umns decreases with increasing cement content. The equation for consolida-
tion time shall be used as a qualiied guess due to the uncertainties of the
permeability in the columns. Long-term settlements and creep settlements
have been discussed in numerous publications; however, this area needs more
investigation together with studies of the increasing strength over time in
DSM. Therefore, creep settlements in DSM columns are left out of the design
today.

10.4.3 Testing and quality control

Testing and quality control of the performance is divided into ield test-
ing, which involves pretesting in test areas as well as testing of production
columns, monitoring and instrumentation of the installation process, and
monitoring of performance of the system (improved soil and structure). A
recent review of the execution and quality control of DSM and mass mix-
ing is reported in Larsson (2005).

10.4.3.1 Field test methods

The mechanical properties of the stabilised soil are controlled in situ with
various types of penetration test methods. Sampling of stabilised soil sam-
ples in fresh columns (MOSTAP) or in hardened columns (coring) can also
be performed. The samples can afterwards be tested in the laboratory.
Visual inspection of the homogeneity of columns can be executed through
trial pits and exposure of columns. Samples to analyse the chemical com-
position can also be collected from trial pits or by coring. Extraction of an
entire column and subsequent testing may also be performed using large
split-tube samplers. However, this method is expensive and has only been
used for research projects and special situations. Plate load tests and test
embankment are also used to determine the elasticity module in the column
and the combined compression module in the stabilised soil volume.

10.4.3.1.1 Column penetration tests (KPS or SCPT)

Column penetration tests or the lime column probe is the most common
test method in Scandinavia. The method is a penetration test method using
a vane with a diameter of 400–500 mm (see Figure 10.19).
 Dry soil mixing 473

B
Section A-A
(d-5)
A

B (mm) d (mm)
400 20
500 15
A 600 15
d
φ 36–44 mm

φ 50 mm

Figure 10.19 Column penetration test (the lime column probe). (From Svenska
Geotekniska Förening (Swedish Geotechnical Society). (2000). Lime and lime-
cement columns. Guide for design, construction and control, Report 2. , Linköping,
Sweden (in Swedish).)

The vane is pressed down into the column with a penetration rate of 20
mm/sec and the push-down force is recorded. The method was developed
from the Iskymeter method from the late 1930s and the mid-1950s. The
Iskymeter, which was developed for penetration tests in very soft clays, was
calibrated with respect to vane tests and fall cone tests. A semi-empirical
relation was developed where the undrained shear strength τfu can be evalu-
ated according to Equation 10.21. In 1979 Boman presented a simpliied
equation to evaluate the undrained shear strength in a lime column.
The equation to determine the shear strength in soft soils using the
Iskymeter method:

 1
0.06 × γ × h ×  1 − 
0.092 × P  St 
τ fu = + (10.21)
 2  2
 1 + S  × A  1 + S 
t t

where:
P = penetration force
γ = density of the soil
A = area of the probe
St = sensitivity of the soil.

Boman proposed a simpliied evaluation to evaluate the undrained shear

strength in a lime column.
474 Ground improvement

F
τ fu = (10.22)
N×A
N is a bearing factor equal to 10 (empirical value) for a probe with the
area 100 cm 2 , according to the Swedish guideline SGF (2000). There is,
however, continued discussion about the bearing factor and, according to
EuroSoilStab, a bearing factor of 10 to 15 can be used.
The test method evaluates a mean value of the shear strength along the
column. The method is suitable for columns <10 m due to the risk of the
vane deviating out of the column. This deviation tendency can be overcome
by predrilling a small centre hole in the column. Columns with greater
shear strength than 300 kPa can in many cases be dificult to penetrate.
The test is regarded as a nondestructive method for embankments where
the tested column is surrounded by a great number of other columns and
support for the embankment is assured. However, for structural founda-
tion works, test columns should be used with reduced bearing capacity in
the inal construction.
For evaluation of the test result, the mantle friction along the probe shall be
taken into account. In recent years the method has been improved by attaching
a CPT (Cone Penetration Test) device to measure the penetration resistance
without the friction. In the CPT device an inclinometer has also been added for
verticality control (Forsgren and Ekström, 2002). A minimum of 5%–10% of
the total tested columns shall be taken in unstabilised soil in order to compare
the shear strength and to evaluate the mantle friction along the probe.

10.4.3.1.2 Reverse column penetration test (PORT)

The reverse column penetration test (PORT; Figure 10.20) is a pull-out
test. A vane is installed below the bottom of the column, with a wire up to
ground surface. The vane can be installed at the same time as the columns
or directly after the column has been installed. The vane shall be installed
a minimum 1 m deeper than the column tip.
The vane is pulled out through the column with a penetration rate of
20 mm/sec and the pull-out force is measured. The shear strength of the
column can then be evaluated according to Equation 10.22. The area of
the vane shall be 100 cm 2 and a bearing factor 10 is recommended in the
Swedish guideline SGF (2000). As previously noted, however, EuroSoilStab
recommends a bearing factor of 10 to 15.
The PORT method evaluates the shear strength similar to the column pen-
etration test. With PORT testing there is no theoretical limitation of column
length and columns with shear strength of up to 600 kPa can be tested. For
evaluation of the test result, the mantle friction along the wire shall be taken
into account. A minimum of 5%–10% of the total tested columns shall be
installed only with the wire to evaluate the mantle friction along the wire.
 Dry soil mixing 475

(a)

a
15–20 mm

a a-a

(b)

Figure 10.20 (a) Reverse column penetration test. (From Holmqvist, L. (1992). The lime
column method. Bygg and Teknik, 7–8:40–44, in Swedish.) (b) Picture of
installation of the probe. (Courtesy of LCM.)

10.4.3.1.3 Cone penetration tests (CPT)

Cone penetration testing is frequently used in Norway and Finland and is
a primary test method in other European countries. However, in Sweden
CPT is used as a complementary test method. The method tests a small
area of the column. Therefore, small local weak zones can have a major
inluence on the test result and the apparent shear strength of the column
cross-section may not be representative from CPT tests. When CPT testing
is used, a larger number of tests are recommended, typically 1%–4% of the
installed columns in order to make a statistical evaluation from which a
476 Ground improvement

mean value and standard deviation can be identiied. The shear strength in
a column can be evaluated according to Equation 10.23:

qc − σ v 0
τ fu = (10.23)
Nc

where:
qc = measured cone resistance
σv0 = total overburden stress
Nc = Bearing factor 10 to 25 (according to EuroSoilStab Nc = 10–13)

10.4.3.1.4 Test embankments and plate load tests

Test embankment and plate load tests are suitable methods for evaluat-
ing the elasticity modulus and combined compression modulus Mcomb for
a composite column/soil stabilised zone. Test embankments are expensive
methods and take a long time but are very valuable, especially for large
projects.
Plate load tests can be performed as traditional load tests or as special
compression tests using a plate under the bottom of the column as counter-
force, with a wire to the column top (Baker, 2000).

10.4.3.1.5 Visual inspection

Visual inspection cannot be used as a test method but is valuable in
understanding variation in the product and in the interlocking zone in panels
and grids. Visual inspection can be performed in trial pits down to 2 or
3 m. In the trial pits, it is possible to take soil samples from the columns
and perform chemical analysis in laboratory. While providing qualitative
information, this would not be a reliable test method for DSM columns
(Figure 10.21).

10.4.3.1.6 Other test methods

Other test methods used for DSM columns and mass stabilisation are the
Finnish vane test, core sampling, extraction of whole column, and rock
sounding (or total sounding).
The Finnish vane test (Figure 10.22) was developed from the ordinary
vane test method. The method is suitable in soft columns and in mass stabi-
lised soil to evaluate the shear strength. The method has limitations in stiffer
columns and has a tendency to disturb the stabilised soil during penetration.
The shear strength can be evaluated according to Equation 10.24.

τ fu = N v × Mmax (10.24)
 Dry soil mixing 477

Figure 10.21 Exposed DSM column.

where:
Nv = Vane factor (345)
M max = Maximum torque.

Extraction of the whole column (Figure 10.23) has been used in some
large projects and in research projects. Extraction is performed by a spe-
cial designed casing, pressed down around the column and extracted by a
mobile crane. This is an expensive test and seldom used.
Core sampling has frequently been used as a test method for wet mixing
columns. In Japan core sampling is a standard test method and the samples
are collected after 5 or 6 days. Typical sample equipment is a Denison sam-
pler, double core tube, and triple core tube. Diameters of the sample tubes
are 86–150 mm. For DSM columns, undisturbed core sampling is dificult
to perform because of the risk of cracks in the samples. It is recommended
that the samples be tested with consolidated-undrained triaxial tests.
MOSTAP sampling is a core sampling method and has been used in a
number of projects in the UK. The method uses standard CPT equipment
and consists of a cone and cutting shoe at the base of the sampling tube. The
sampling method is easy to use, and samples of 36–65 mm in diameter can
be obtained. The samples lengths are 1.20 m and are retained within a stock-
ing in a UPVC liner. The liner is sealed within airtight end caps. Samples are
478 Ground improvement

32
3

6
65
20

132

Figure 10.22 Finnish vane test. (From Halkola, H. (1983). In-situ investigation of deep
stabilized soil. Proceedings of the 8th European Conference on Soil Mechanics
and Foundation Engineering, May 23–26, 1983.)

preferably taken in fresh columns (soon after installation) and stored in labo-
ratory for hardening. After hardening the sample tube is split and laboratory
tests can be performed on the samples. This method has shown good results
especially for visual inspection as well as for chemical testing.
Soil/rock sounding and total sounding are common Swedish and
Norwegian methods. These methods are commonly used for predrilling
a centre hole in the columns before the column penetration tests are per-
formed. The penetration resistance can be roughly estimated by adding a
bearing capacity or a correlation factor to the results from column pen-
etration tests, which identiies the undrained shear strength. However, this
method alone is not reliable.

10.4.3.2 Performance of penetration tests

The performance of the testing is crucial to the outcome of the test results.
Here is a simple guideline for test procedure:

(1) Exposure of column head and survey of the column position and level.
(2) Documentation of the column head (e.g., photo documentation).
(3) Predrilling with a soil/rock sounding. During drilling, registration
of penetration force, rotation speed (if necessary), and torque. Spoil
water shall not be used.
(4) Execute column penetration test, with or without the CPT device.
(5) Taking short notes (e.g., if the probe tends to deviate out of the col-
umn) or other observations that could be of interest to the designer.

The test result is best presented graphically for each column as well as for
the penetration force of the predrilling. All the tests from one test session
 Dry soil mixing 479

Figure 10.23 Extracted column. (From Axelsson, M. (2001). Djupstabilisering med

Kalkcementpelare-Metoder for Produktionsmassig Kvalitetskontroll i Falt (Deep Stabilization
with Lime Cement Columns – Methods for Quality Control in the Field), Report No. 8.
Linkoping, Sweden: Swedish Deep Stabilization Research Centre.)

should be summarised and presented graphically with average value,

median value and bound values.
Example requirements for ield tests with the column penetration method are
shown in Table 10.3. Minimum required shear strength and average required
shear strength at different levels and time scales are valuable information.
480 Ground improvement

Table 10.3 Example of a test table for evaluation of undrained shear

strength with the column penetration test method
Depth below Average shear Lower 15% Minimum strength
working platform strength (kPa) percentile (kPa) of local value (kPa)
0–0.5 - - -
0.5 ≥100 ≥50 -
0.5–2.0 Straight line Straight line -
interpolation interpolation
>2.0 ≥150 ≥120 90

If the shear strength is not fulilling the requirements, then additional

testing in a nearby location should be the next step.
Data from the penetration force is commonly presented as a loating aver-
age of z − 200 mm to z + 200 mm. This gives an average of 400 mm. The
loating average is today practiced in many projects in order to reduce the
inluence of locally large variations in the columns. For soil improvement
with DSM columns, a locally weaker or stiffer zone <0.5 m has no signii-
cant effect on the construction, hence the boundary of the loating average.

10.5 CASE HISTORIES

Dry soil mixing also referred to as the Nordic method was rarely used
outside the Nordic countries until the late 1990s. After this the number of
projects has signiicantly increased and today the method is well accepted
and used successfully in numerous countries around the world. Four case
histories from different countries are presented in this chapter.

10.5.1 Thames estuary, UK

In the UK, a large number of projects have been carried out during the last
few years. The projects include DSM improvement for roads, increasing of
stability in harbour areas (Lawson et al. 2005), improvement of stability
for railway embankments (Pye et al. 2012a), land reclamation, mass mix-
ing, and support for deep excavations.
Soil improvement by using DSM columns (Figure 10.24) was carried out
to provide a renewed access road for a large new ‘Energy from Waste’ facil-
ity near London on the Thames estuary. An existing route to the new plant
was sited over soft alluvial deposits and was in very poor condition and quite
unsuitable for traficking by trucks during construction, as well as for the
future operation of the new plant. The soil proile underneath the existing
road was soft clay overlying 2–3 m of peats overlying very soft clay with
peat layers in the clay. The soft deposit was 6–9 m deep overlying a dense
sand deposit (see Figure 10.25). The moisture content varied between 160%–
400% with an undrained shear strength of 6–21 kPa in the peat layer.
 Dry soil mixing 481

DSM columns were selected to increase the stiffness of the soft soil in
order to reduce settlements and provide a new foundation for the access
road. DSM columns with diameters of 800 mm were selected. These
were formed in a rectangular grid with 1.1 m spacing, giving a replace-
ment ratio of 0.41. The selected binder was Cement CEM I with a dosage
rate of 200 kg/m3. The amount and type of binder to be used was based
upon evaluation of the soil characteristics from the site investigations
together with the geotechnical engineer’s local knowledge of soil mix-
ing. Soil samples from the site were collected for laboratory mixing tests
in order to verify the stabilisation effect. A pretest was conducted to
demonstrate the performance of the improvement. Column penetration
testing was carried out after 5–7 days and 14–16 days according to the
ield test scheme (Figures 10.25 and 10.26). The preselected binder and
amount of binder demonstrated good increase in shear strength, and the
design shear strength of 150 kPa (28 days’ strength) was demonstrated
within 7 days test.
Installation of production columns started after inalising the pretest,
which demonstrated the performance of the DSM columns, with binder
content of 100 kg/m column (200 kg/m3) and mixing energy of 350 BRN.
In total, some 7,500 columns were installed at an average depth of 7.8 m.
Testing was performed using CPT, column penetration tests (KPS tests)
and reverse column penetration tests (pull-out test, FOPS). The column
penetration tests showed the most reliable test results and were selected
as the main test method. Evaluation of column shear strength at 7–14
days testing showed a wide range of strength. The evaluated average shear
strength was 2–3 times design shear strength.
This case study demonstrates that DSM columns can be used in peaty
soils with high organic content and high moisture content, which report-
edly are dificult to improve.

Figure 10.24 Installation of DSM. (Courtesy of Keller Géotechnique.)

482 Ground improvement

Figure 10.25 Field test with column penetration tests. (Courtesy of Keller Géotechnique.)

Figure 10.26 Exposed columns for visual examination. (Courtesy of Keller Géotechnique.)

10.5.2 Jewfish Creek, Key Largo, Florida

In the United States, improvement by DSM columns and mass mixing has
been used since the early 1990s. However, in the last decade there has
been a major increase in the use of soil improvement techniques, especially
 Dry soil mixing 483

Table 10.4 Jewish Creek project-speciic information

Source Hayward Baker, Mann, J.A, Sehn, A., Burk, G.
Location Highway between Florida City and Key West in Key Largo,
Florida, USA
Construction site 8 km (5 miles) long, 12 m (40 ft) wide (see Figure 10.27)
Soils Very soft organic mangrove with peat, muck, and marl,
organic content 40%–60%, moisture content w = 50%–650%
Design requirements Design in-place minimum shear strength of 75 kPa (1,500 lbs/
ft2). Long-term settlement maximum 50 mm in 5 years.
Applied DM method Mass mixing with rotary head shaft (see Figure 10.28)
Installation data Volume to be mixed 270,000 m3 (350,000 yd3)
Design shear strength 250 kPa under the rails, 150 kPa remaining area
Binder type and factor Slag 75% and cement 25%, 140–160 kg/m3
Field test methods Sampling by coring (see Figure 10.29), Penetration test by
KPS-vane, PORT, test embankment (see Figure 10.30)

for supporting lood levees and loodwalls after Hurricane Katrina and
Hurricane Rita devastated parts of the southern United States.
The project Jewish Creek is located in Key Largo, Florida (Table
10.4). The existing road was extended by 12m in width and 8km in
length, see Figure 10.27. A typical mass-mixing setup is illustrated in

Figure 10.27 Site location of the Jewish creek project.

484 Ground improvement

Figure 10.28. Field test methods are illustrated in Figures 10.29 and
10.30. As an alternative to traditional solutions, such as removing and
replacing the soft mangrove material, surcharge or installation of piles
with structural platform in-situ treatment with mass mixing (Figure
10.31) was selected.
Conversion between psi and kPa, 1psi ≈ 6.89 kPa

(a) (b)

Figure 10.28 Mass mixing equipment. (a) Mixing rig. (b) Blending station. (Courtesy of
Hayward Baker.)

(a)

(b) 20

15
Relative frequency (%)

10

0
0 50 100 150 200 250 300
Unconfined compressive strength (psi)

Figure 10.29 (a) Typical core samples. (b) Compression strength from core samples
taken. (Courtesy of Hayward Baker.)
 Dry soil mixing 485

(a)

Date
(b)
12/26/05

1/16/06

1/23/06

1/30/06

2/13/06

2/20/06

2/27/06
1/2/06

1/9/06

2/6/06

3/6/06
0.00 30

0.05 25

0.10 20
Settlement (ft)

Fill height (ft)

0.15 15

0.20 10

0.25 N. Plate A C. Plate B 5

S. Plate C Fill height
0.30 0

Figure 10.30 (a) Test embankment. (b) Settlement measurements. (Courtesy of

Hayward Baker.)

Figure 10.31 Production of mass mixing on site. (Courtesy of Hayward Baker.)

10.5.3 Road 45 and Norway/Väner Route, Sweden

Road 45 and Norway/Väner route is one of the largest infrastructure
projects in Sweden, which started in 2007 and will be inalised in 2012
(see Table 10.5). The project expands the European Road E45 between
Gothenburg and Trollhättan (80 km) from a two-lane to a four-lane high-
way. The railway system is expanded from an extreme one-lane track to a
two-lane high-speed railway (train speed 250 km/h).
486 Ground improvement

Table 10.5 Project-speciic information for road 45 and Norway/Väner route

Source Swedish Road Authorities, www.banavag.se
Location Road- and railway (high-speed trains) between
Gothenburg and Trollhättan, Sweden.
Construction site 80 km highway and railway (see Figure 10.32)
Soils Very soft clay to large deposit more than 100 m locally.
Typical undrained shear strength 8–25 kPa, moisture
content w = 40%–150%, sensitivity up to St < 400.
Design Design undrained shear strength of 150 kPa. Long-term
requirements in settlement maximum 250 mm in 40 years, stability;
DSM columns factor of safety 1.5–1.65 (1.65 in quick clay areas).
Applied DM Lime/cement columns with diameter 600 mm
method
Installation data Approximately 9.2 million linear metres installed up to
25-m depth. Interlocking columns in panels and grids for
stability and high-speed trains
Binder type and Lime 50% and cement 50%, 90–30 kg/m3, some areas lime
factor 33%, cement 33%, and ly ash 33%.
Field test methods Test embankments, Álen et al. (2005b), column
penetration tests with CPT devise and inclinometer
device.

Location of the project is in the Göta Älv River Valley (see Figure 10.32).
The geological formation in the valley is characterized by large clay depos-
its with highly sensitive clays (quick clays) overlying bedrock. Slope stabil-
ity is a major issue along the Göta Älv River, and two major slip failures
with human loss have occurred in the area (the Surte failure 1950 and the
Göta failure 1957). In addition, there have been a large number of failures
resulting in material losses.
The soft clay deposits are stabilised by approximately 9 million linear
DSM columns (Figure 10.33). Columns are installed to increase stability,
reduce settlements, and prevent mitigation of ground vibrations connected
to high-speed trains. Other applications are foundations of structures such
as bridges (see Figures 10.14a, b and 10.15).
Before the project started (5–7 years), test embankments were installed
and monitored. Field and laboratory tests were performed in order to
determine binder combination, amount of binder, and mixing param-
eters for the total project before construction started. This is, however, an
unusual way of designing because all contractors are restricted to certain
parameters. In construction, pretesting is performed for each part of the
project. Pretesting of DSM columns is carried out in test areas to verify the
predesigned installation concept. Field testing is performed on two occa-
sions, 12–16 days and 28–34 days. No other time periods are allowed for
testing.
 Dry soil mixing 487

Uddevalla
Railway
Highway 44
Trollhättan

Stenröset
Velanda
Torpa

E15
Upphärad

Lilla Edet Prässebo

Göta Kärra
Stenungsund
Torbacken
Nygärd
Lödöse
Hede
E6
Göta älv Alvhem

Alvängen
Alingsas
Nol
Kungälv
Nödinge

Bohus

Surte
Agnesberg
E20
Marieholm

Olskroken

Göteborg 40

Figure 10.32 Site map of the project. (From Banaväg in Väst website, http://www.
traikverket.se/banavag.)

(a) (b)
PEAB Förprovning 12-16 dygn
Kontrollobjekt 7
Sammanfattande diagram 30 kg/m, 20 mm/r
Sektion 000+000 Tvärmatt: ±Xm

Skjuvhällfasthet [kPa]
0 50 100 150 200 250 300 350 400 450
0 Cu förväntad
undre fraktil 15%
Cu förväntat
1 medel
Cu, medel
Cu, undre fraktil 15%
2 520C105
520C109
3 520C113
520C117
4 520C121
Djup under arbetsysta [m]

524C54
5 524C106
524C107
524C111
6
524C114
524C115
7
524C118
524C119
8
524C122
524C123
9

10

11

12

Tillverkningsdatum: 2009- 12-17 Verktyg: PB600 Stigning: 200 mm/r

Provdatum: 2009-12-29 Faktor: 12,5 Diameter: 600 mm
Projekt nr: 2008-24 3 Mix: 30 kg/m
Pelarlängd: 12 m, 20 m Sond: 500 × 15 mm

Figure 10.33 (a) Testing of DSM columns. (b) Printout of KPS tests. (Courtesy of dmixab.)
488 Ground improvement

14.9 m
New 2.5 m 2.5 m Rehab.
track track
Working
Embankment platform

22.9 m

Figure 10.34 Typical cross-section of the railway embankment and treated zone. (From
Raju, V. R., Abdullah, A. and Arulrajah, A. (2003). Ground treatment using
dry deep soil mixing for a railway embankment in Malaysia, Proceedings
of the 2nd International Conference on Advances in Soft Soil Engineering and
Technology, July 2–4, 2003, Putrajaya, Malaysia.)

10.5.4 Railway embankment in Malaysia

The inal case history concerns the provision of foundations for a railway
embankment along a half-mile section of high-speed line between Rawang and
Ipoh in Malaysia (Figure 10.34). Ground conditions consisted of very soft allu-
vial deposits (Figure 10.35 and Table 10.6), and dry soil mixing was selected
as an effective means of providing the design requirements (Raju et al. 2003).
Post treatment testing included an area loading test which provided excellent
veriication that the speciied performance was achieved (Figure 10.36).

(a) (b)
0 2 4 6 8
0
qs [MPa]
2

10

12
z [m]

Figure 10.35 (a) Typical CPT log. (b) The LCM machine at work. (From Raju, V. R.,
Abdullah, A. and Arulrajah, A. (2003). Ground treatment using dry deep
soil mixing for a railway embankment in Malaysia, Proceedings of the 2nd
International Conference on Advances in Soft Soil Engineering and Technology,
July 2–4, 2003, Putrajaya, Malaysia.)
 Dry soil mixing 489

Table 10.6 Railway embankment in Malaysia

Source Raju, Abdullah, and Arulrajah (2003)
Location Railway line between Rawang and Ipoh, Malaysia
Construction site 800 m long, 20–25 m wide (see Figure 10.34)
Soils Very soft silty clay or clayey silt to loose silty clayey sand,
typical CPT log see Figure 10.35a, moisture content w =
50%– 70%, groundwater ca. 1 m below ground surface
Embankment height and 1.5–3 m, equivalent trafic load 30 kPa
load
Design requirements Train speed 160 km/h, max. settlement 25 mm in 6
months of operation, max. differential settlement 0.1%
along the centreline, safety factor for slope failure
1.5 (long term)
Applied DM method (Lime-cement column), single shaft (Figure 10.35)
Column data Diameter 0.6 m, length 7–14 m, overall 50,000 lin. m
Column pattern Detached columns, square/rectangle, 1–1.3 m c/c under
the rails (ap = 28 to 17%), 1.4–1.5 m c/c remaining area
(ap = 14 to 13%)
Design shear strength 250 kPa under the rails, 150 kPa remaining area
Binder type and factor Portland cement 100%, 100–150 kg/m3
Embankment Geotextile 100/50 kN/m (longitudinal/transverse
reinforcement direction)
Observed performance Settlement below 10 mm for embankment 1–1.5 m,
lateral displacement below 15 mm, loading test
(see Figure 10.36).

Kentledge blocks

0 200 400 600 800 1000 1200

0 Grillage beams
Load [kN] Concrete
2 Sand block

6
s [mm]
8
Load cycle-1 Load cycle-2

Figure 10.36 Control static loading test over an area of 3 × 3 m, 4 columns. (From
Raju, V. R., Abdullah, A. and Arulrajah, A. (2003). Ground treatment using dry
deep soil mixing for a railway embankment in Malaysia, Proceedings of the 2nd
International Conference on Advances in Soft Soil Engineering and Technology, July
2–4, 2003, Putrajaya, Malaysia.)
490 Ground improvement

REFERENCES

Åhnberg, H. and Holm, G. (1986). Kalkpelarmetoden – Resultat av 10 Års Forskning

och Praktisk Användning samt Framtida Utveckling (The Lime Column Method –
Results from Research and Practical Applications during 10 Years and Future
Development), Report No. 31. Linköping, Sweden: Swedish Geotechnical
Institute.
Åhnberg, H., Johansson, S.E., Retelius, A., Ljungkrantz, C., Holmqvist, L., and
Holm, G. (1995). Cement och Lalk för Djupstabilisering av Jord – En Kemisk
Fysikalisk Studie av Stabiliseringseffekter (Cement and Lime for Stabilisation
of Soil at Depth – a Chemical Physical Investigation of Soil Improvement
Effects), Report No. 48. Linköping, Sweden: Swedish Geotechnical Institute.
Åhnberg, H., Bengtsson, P-E. and Holm, G. (2001). Effect of initial loading on the
strength of stabilised peat. Ground Improvement, 7(1):9–23.
Åhnberg, H. (2006). Strength of Stabilized Soils – A Laboratory Study on Clays
and Organic Soils Stabilised with Different Types of Binder. PhD Thesis, Lund
University, Sweden.
Álen C., Baker, S., Bengtsson, P-E., and Sällfors, G. (2005a). Lime/cement column
stabilised soil – a new model for settlement calculation. Proceedings from the
International Conference on Deep Mixing Best Practice and Recent Advances,
May 23–25, 2005, Stockholm, Sweden, pp. 205–212.
Álen C., Baker, S., Ekström, J., Hallingberg, A., Svahn, V. and Sällfors, G. (2005b).
Test embankments on lime/cement stabilized clay. Proceedings from the
International Conference on Deep Mixing Best Practice and Recent Advances,
May 23–25, 2005, Stockholm, Sweden, pp. 213–220.
Axelsson, M. (2001). Djupstabilisering med Kalkcementpelare-Metoder för
Produktionsmässig Kvalitétskontroll i Fält (Deep Stabilization with Lime
Cement Columns – Methods for Quality Control in the Field), Report No. 8.
Linköping, Sweden: Swedish Deep Stabilization Research Centre.
Axelsson, M. and Larsson S. (2003). Column penetration tests for lime-cement col-
umns in deep mixing – experiences in Sweden, Proceedings from the Third
International Conference on Grouting and Ground Treatment, February
10–12, 2003, New Orleans, Louisiana, United States, ASCE Geotechnical
Special Publication No.120, pp. 681–694.
Babasaki, R., Terashi, M., Suzuki, T., Maekawa, A., Kawamura, M., and Fukazawa, E.
(1996). Japanese Geotechnical Society Technical Committee Report: Factors
inluencing the strength of improved soil. In: Yonekura R., Terashi M., and
Shibazaki M. (eds.), Grouting and Deep Mixing: Proceedings of the 2nd
International Conference on Ground Improvement Geosystems, May 14–17,
1996, Tokyo, Japan: Balkema Publishers, pp. 913–918.
Baker, S. (2000). Deformation Behavior of Lime/Cement Column Stabilized Clay,
Report No. 7. Linköping, Sweden: Swedish Deep Stabilization Research Centre.
Bengtsson, P-E. and Karlsson, M. (2006). Dynamisk Påverkan, Working Report No.
39. Linköping, Sweden: Swedish Deep Stabilization Research Centre.
Boman, P. and Broms, B. (1975). Stabilization of cohesive soil by lime col-
umns. Proceedings of the Nordic Geotechnical Meeting, May 22–24, 1975
Copenhagen, Denmark, pp. 265–279 (in Swedish).
 Dry soil mixing 491

Boman, P. (1979). Quality Control of Lime Column. Part 2: Results from Investigations
at Sollentunaholm. Stockholm, Sweden: Royal Institute of Technology (in
Swedish).
Bredenberg, H. (1999). Keynote lecture: equipment for deep soil mixing with the dry
jet mix method. Proceedings of the International Conference on Dry Methods for
Deep Soil Stabilization, October 13–15, 1999, Stockholm, Sweden, pp. 323–331.
Bredenberg, H., Holm, G., and Broms, B., eds. (1999). Dry Mix Methods for
Deep Soil Stabilization: Proceedings of the International Conference on
Dry Methods for Deep Soil Stabilization, October 13–15, 1999, Stockholm,
Sweden.
Broms, B. (1984). Stabilization of Soil with Lime Columns – Design Handbook, 3rd
ed. Kungsbacka, Sweden: Lime Column Method AB.
Broms, B. (1999). Keynote lecture: Design of lime, lime/cement and cement columns.
Proceedings of the International Conference on Dry Methods for Deep Soil
Stabilization, October 13–15, 1999, Stockholm, Sweden, pp. 125–154.
Dahlström, M. and Eriksson, H. (2005). Mass stabilisation in Smista Allé (2002)
and Moraberg (2003) by cell and block stabilisation methods. Proceedings
from the International Conference on Deep Mixing Best Practice and Recent
Advances, May 23–25, 2005, Stockholm, Sweden, pp. 425–429.
Dahlström, M. and Wiberg, D. (2012). Dry soil mixing and vibro replacement in com-
bination for a high embankment. Proceedings from the International Conference
on Grouting and Deep Mixing, February 16–18, 2012, New Orleans, Louisiana,
United States.
EuroSoilStab. (2002). Development of Design and Construction Methods to Stabilise
Soft Organic Soils: Design Guide Soft Soil Stabilisation, CT97-035I, European
Commission Project BE 96-3177.
EN14679:2005 (2005). European Standard, Execution of Special Geotechnical
Works – Deep Mixing. Publisher information
Filz, G., Adams, T., Navin, M., and Tempelton, A.E. (2012) Keynote lecture: Design
of deep mixing for support of levees and loodwalls. Proceedings from the
International Conference on Grouting and Deep Mixing, February 16–18,
2012, New Orleans, Louisiana, United States.
Forsgren, I. and Ekström, J. (2002–2012). Personal communication.
Forsman, J., Maijala, A., and Järvinen, K. (2008). Case stories, harbours – mass
stabilisation of contaminated dredging mud in Sörnäinen, Helsinki. Proceedings
of the International Mass Stabilisation Conference 2008, October 8–10, 2008,
Lahti, Finland.
Halkola, H. (1983). In-situ investigation of deep stabilized soil. Proceedings of the
8th European Conference on Soil Mechanics and Foundation Engineering,
May 23–26, 1983, Helsinki, Finland.
Halkola, H. (1999). Keynote lecture: quality control for dry mix method. Proceedings
of the International Conference on Dry Methods for Deep Soil Stabilization,
October 13–15, 1999, Stockholm, Sweden, pp. 285–294.
Hansbo, S. (1979). Consolidation of clay by bandshaped prefabricated drains,
Ground Engineering 12(5): 21–25.
Hayashi, H., Nishikawa, J., Ohishi, K. and Terashi, M. (2003). Field Observation
of Long-Term Strength of Cement Treated Soil, ASCE Geotechnical Special
Publication No. 120, pp. 598–609.
492 Ground improvement

Hebib, S. and Farrell, E.R. (2003). Some experience on the stabilisation or Irish
peats, Canadian Geotechnical Journal 40:107–120.
Holm, G. (1999). Keynote lecture: applications of dry mix methods for deep soil
stabilization. Proceedings of the International Conference on Dry Methods for
Deep Soil Stabilization, October 13–15, 1999, Stockholm, Sweden, pp. 3–14.
Holm, G. (2002). Nordic dry deep mixing method – execution procedure, Proceedings
of the Deep Mixing Workshop 2002, Port and Airport Research Institute &
Coastal Development Institute of Technology, Tokyo, Japan.
Holm, G., Andréasson, B., Bengtsson, P-E., Bodare, A., and Eriksson, H. (2002).
Mitigation of Track and Ground Vibrations by High Speed Trains at Ledsgård,
Sweden, Report No. 10. Linköping, Sweden: Swedish Deep Stabilization
Research Centre.
Holmqvist, L. (1992). The lime column method. Bygg and Teknik, 7–8:40–44 (in
Swedish).
Horpibulsuk, S., Miura, N. and Nagara, T. (2003). Assesment of strength develop-
ment in cement-admixed high water content clay with Abrams’ law as a basis.
Géotechnique 53(4): pp. 439–444.
Janz, M. and Johansson, S.E. (2002). The function of different binding agents in deep
stabilization, Report No. 9. Linköping, Sweden: Swedish Deep Stabilization
Research Centre.
Jelisic, N. and Leppännen, M. (1999). Mass stabilization of peat in road and railroad
construction. Proceedings of the International Conference on Dry Methods for
Deep Soil Stabilization, October 13–15, 1999, Stockholm, Sweden, pp. 59–64.
Jelisic, N. and Leppännen, M. (2005). Remediation of contaminated land
of Sörnäinen, Helsinki, by using mass stabilization. Proceedings of the
International Conference on Deep Mixing Best Practice and Recent Advances,
May 23–25, 2005, Stockholm, Sweden, pp. 353–356.
Kitazume, M., Tabata, T., Ishiyama, S. and Ishikawa, Y. (1996a). Model tests on
failure pattern of cement treated retaining wall, In: Yonekura R., Terashi M.,
and Shibazaki M. (eds.) Grouting and Deep Mixing: Proceedings of the 2nd
International Conference on Ground Improvement Geosystems, May 14–17,
1996, Tokyo, Japan, pp. 509–514.
Kivilö, M. and Broms, B. (1999). Mechanical behavior and shear resistance of lime/
cement columns. Proceedings of the International Conference on Dry Methods for
Deep Soil Stabilization, October 13–15, 1999, Stockholm, Sweden, pp. 193–200.
Lawson, C.H, Spink, T.W, Crawshaw, J.S. and Essler R.D. (2005). Veriication
of dry soil mixing at Port of Tilbury, UK. Proceedings of the International
Conference on Deep Mixing Best Practice and Recent Advances, May 23–25,
2005, Stockholm, Sweden, pp. 453–462.
Larsson, S. (1999). The mixing process at the dry jet mixing method. Proceedings
of the International Conference on Dry Methods for Deep Soil Stabilization,
October 13–15, 1999, Stockholm, Sweden, pp. 339–346.
Larsson, S. (2003). Mixing Process for Ground Improvement by Deep Mixing,
Doctoral Thesis, Royal Institute of Technology, Stockholm, Sweden.
Larsson, S. (2005). Keynote lecture: state of practice report – execution, monitor-
ing and quality control. Proceedings of the International Conference on Deep
Mixing Best Practice and Recent Advances, May 23–25, 2005, Stockholm,
Sweden, pp. 732–785.
 Dry soil mixing 493

Larsson, S., Dahlström, M. and Nilsson, B. (2005a). Uniformity of lime-cement col-

umns for deep mixing: a ield study. Ground Improvement 9(1):1–15.
Larsson, S., Stille, H. and Olsson, L. (2005b). On horizontal vaiability in lime-cement
columns in deep mixing. Géotechnique 55(1):33–44.
Larsson, S., Dahlström, M. and Nilsson, B. (2005c). A complementary ield test on
the uniformity of lime-cement columns. Ground Improvement 9(2): 67–77.
Löfroth, H. (2005). Properties of 10-year-old lime-cement columns, Proceedings
of the International Conference on Deep Mixing Best Practice and Recent
Advances, May 23–25, 2005, Stockholm, Sweden, pp. 119–127.
Martin, G.R., Arulmoli, K., Yan, L., Esrig, M.I., and Capelli, R.P. (1999). Dry mix soil-
cement walls: An application for mitigation of earthquake ground deformations
in soft or liqueiable soils, Proceedings of the International Conference on Dry
Methods for Deep Soil Stabilization, October 13–15, 1999, Stockholm, Sweden,
pp. 37–44.
Muro, T., Fukagawa, R. and Mukaihata, K. (1987a). Fundamental study of mixing
conditions affecting the mechanical properties of soil-cement piles. Doboku
Gakkai Ronbunshu 385:98–195 (in Japanese).
Muro, T., Fukagawa, R. and Mukaihata, K. (1987b). Effect of mixing condition
on deformation and strength characteristics of cement treated soil. Ehime
Daigaku Kogakubu Kioy 11(2):393–403 (in Japanese).
Nagaraj, T.S., Muira, N. Yaligar, P.P. and Yamadera, A. (1996). Predicting strength
development by cement admixture based on water content. In: Yonekura R.,
Terashi M., and Shibazaki M. (eds.), Grouting and Deep Mixing: Proceedings
of the 2nd International Conference on Ground Improvement Geosystems,
May 14–17, 1996, Tokyo, Japan, pp. 431–436.
Nishida, K., Koga, Y., and Muira, N. (1996). Energy consideration of dry jet mixing
method. In: Yonekura R., Terashi M., and Shibazaki M. (eds.), Grouting and
Deep Mixing: Proceedings of the 2nd International Conference on Ground
Improvement Geosystems, May 14–17, 1996, Tokyo, Japan, pp. 643–648.
Porbaha, A., Shibuya, S. and Kishida, T. (2000). State of the art in deep mixing
technology. Part III geomaterial characterization. Ground Improvement
4(3):91–110.
Pye, N., O´Brien, A., Essler, R.D. and Adams, D. (2012a). Deep dry soil mixing to sta-
bilize a live railway embankment across Thrandestone Bog. Proceedings from
the International Conference on Grouting and Deep Mixing, February 16–18,
2012, New Orleans, Louisiana, United States.
Pye, N., O´Brien, A., Essler, R.D. and Adams, D. (2012b). Laboratory and ield test-
ing for deep dry soil mixing to stabilize a live railway embankment across
Thrandestone Bog. Proceedings from the International Conference on
Grouting and Deep Mixing, February 16–18, 2012, New Orleans, Louisiana,
United States.
Raju,V. R., Abdullah, A. and Arulrajah, A. (2003). Ground treatment using dry
deep soil mixing for a railway embankment in Malaysia, Proceedings of
the 2nd International Conference on Advances in Soft Soil Engineering and
Technology, July 2–4, 2003, Putrajaya, Malaysia.
Rogers, C.D.F. and Glendinning, S. (1996). The role of lime migration in lime
pile stabilization of slopes. Quarterly Journal of Engineering Geology
29:273–284.
494 Ground improvement

Rogers, C.D.F., Glendinning, S. and Holt, C.C. (2000a). Slope stabilization using
lime piles – a case study, Ground Improvement 4(4):165–176.
Rogers, C.D.F., Glendinning, S. and Troughton, V.M. (2000b). The use of additives
to enhance the performance of lime piles. Proceedings of the 4th International
Conference on Ground Improvement Geosystems, June 7–9, 2000, Helsinki,
Finland, pp. 127–134.
Svenska Geotekniska Förening (Swedish Geotechnical Society). (2000). Lime and
lime cement columns. Guide for design, construction and control, Report 2,
Linköping, Sweden (in Swedish).
Terashi, M. (2003). The State of Practice in Deep Mixing Methods, ASCE
Geotechnical Special Publication No. 120, pp. 25–49.
Terashi, M. (2005). Keynote lecture: design of deep mixing in infrastructure applica-
tions. Proceedings of the International Conference on Deep Mixing Best Practice
and Recent Advances, May 23–25, 2005, Stockholm, Sweden, pp. K25–K45.
Timoney, M.J., McCabe B.A., and Bell, A.L. (2012). Experiences of dry soil mixing in
highly organic soils, Proceedings of the Institution of Civil Engineers – Ground
Improvement 165(1):3–14.
Yoshizawa, H., Okumura, R., Hosya, Y., Sumi, M. and Yamada, T. (1997). Japanese
Geotechnical Society Technical Committee Report: factors affecting the qual-
ity of treated soil during execution of DMM. In: Yonekura R., Terashi M.,
and Shibazaki M. (eds.) Grouting and Deep Mixing: Proceedings of the 2nd
International Conference on Ground Improvement Geosystems, May 14–17,
1996, Tokyo, Japan, pp. 931–937.