Report 3

Stabilization of Organic Soils by
Cement and Puzzolanic Reactions
– F

EASIBILITY

TUDY

Karin Axelsson
Sven-Erik Johansson
Ronny Andersson

Svensk Djupstabilisering

Swedish Deep Stabilization Research Centre

English
Translation
July 2002

Swedish Deep Stabilization Research Centre

The Swedish Deep Stabilization Research Centre coordinates research and development activities in
deep stabilization of soft soils with lime-cement columns. A joint research programme based on the
needs stated by the authorities and the industry is being conducted during the period 1996 – 2004.
Members of the Centre include authorities, lime and cement manufacturers, contractors, consultants,
research institutes and universities.

The work of the Swedish Deep Stabilization Research Centre is financed by its members and by
research grants.

The Swedish Deep Stabilization Research Centre is located at the Swedish Geotechnical Institute and
has a Steering Committee with representatives choosen from among its members.

Further information on the Swedish Deep Stabilization Research Centre can be obtained from the
Project Manager, Mr G Holm, tel: +46 13 20 18 61, +46 70 521 09 39, fax: +46 13 20 19 14 or
e-mail: goran.holm@swedgeo.se, http://www.swedgeo.se/sd.htm.

Report

Stabilization of Organic Soils

by Cement and

Puzzolanic Reactions

EASIBILITY

TUDY

Karin Axelsson

Sven-Erik Johansson

Ronny Andersson

Svensk Djupstabilisering

Swedish Deep Stabilization Research Centre

English Translation July 2002

Swedish Deep Stabilization Research Centre, Report 3

Swedish Deep Stabilization Research Centre
c/o Swedish Geotechnical Institute
SE-581 93 Linkoping

Phone: +46 13 20 18 62
Fax: 013-20 19 13
E-mail: birgitta.sahlin@swedgeo.se

1402-2036
SD-R--00/3--SE

Report

ISSN

ISRN

Stabilization of Organic Soils

Foreword – English Translation

Translation of this document was made possible through collaborative effort of Swedish Deep Sta-
bilization Research Centre and the US National Deep Mixing (NDM) program. The mission of both
organizations is to facilitate advancement and implementation of deep mixing technology through
partnered research and dissemination of international experience.

The Swedish Deep Stabilization Research Center with its headquarter at the Swedish Geotechnical
Institute (SGI) coordinates research primarily on dry mix methods. An international conference on
“dry mix methods for deep soil stabilization” was held in Stockholm in 1999. The findings from
research and state-of-the practice in Europe, particularly on stabilization of organic soil, were pre-
sented and published in the proceedings. SGI is also a partner of EuroSoilStab, the R&D project
focusing on organic soils and infrastructure applications.

The National Deep Mixing program coordinates a program of deep mixing research in the US. The
international workshop on “Deep mixing technology for infrastructure development” was held in
Oakland, California in 2001. A forum of users and experts from industry, government and academia
examined the current practice and research needs; and identified the challenges ahead for imple-
mentation of deep mixing technology.

We hope dissemination of international experience serves as a step toward our better understanding
and promotion of this innovative technology in the construction industry.

July 2002

Göran Holm, Linkoping, Sweden

Ali Porbaha, California, USA

Swedish Deep Stabilization Research Centre, Report 3

Foreword

The present feasibility study, “Stabilization of Organic Soils by Cement and Pozzolanic Reactions”,
was carried out as a project at the Swedish Deep Stabilization Research Centre (SD) in summer
1996 and reports the state of our knowledge at that time. The findings have since been actively ap-
plied at the Centre and within the EU project EuroSoilStab. The work comprises an investigation of
the strength of stabilized mud and peat. The report has served as a basis for SD’s continued study of
the properties of stabilized earth—relevant parameters.

The work of the feasibility study is divided into laboratory trials, evaluation, and comparison with
previously reported projects. The trials were done on mud and peat from two locations, Arlanda
airport and the Örebro–Arboga motorway, where reinforcement works including deep stabilization
were in progress at the time.

The authors wish to thank Helen Åhnberg of the Swedish Geotechnical Institute and Elina Parkkin-
en of Lohja Rudus Oy AB for their assistance with the work.

The report was edited by Jan Lindgren of the Swedish Geotechnical Institute.

Danderyd (Stockholm), 11 August 2000

Björn Paulsson

Karin Axelsson Sven-Erik Johansson Ronny Andersson

Chair, SD

Stabilization of Organic Soils

Contents

Foreword

............................................................................................................................. 3

Foreword English Translation

.................................................................................................... 4

Summary

............................................................................................................................. 7

Introduction

.......................................................................................................................... 9

1.1

Purpose ........................................................................................................................... 9

1.2

Background .................................................................................................................... 9

Mass stabilization

.............................................................................................................. 10

Stabilization trials in mud and peat – execution

....................................................... 12

3.1

Classification and properties of mud and peat ............................................................. 12

3.2

Pore water analysis ...................................................................................................... 13

3.3

Stabilizing agents ......................................................................................................... 14

3.4

Test programme ........................................................................................................... 14

3.5

Sample preparation and testing .................................................................................... 15

The influence of various factors on stabilization effectiveness

........................... 17

4.1

Effect of stabilizers ...................................................................................................... 17
4.1.1 Cement and lime ................................................................................................ 18
4.1.2 Fly ash ............................................................................................................... 19
4.1.3 Granulated blast furnace slag ............................................................................ 19
4.1.4 Filler materials ................................................................................................... 20

4.2

stabilizer quantities ...................................................................................................... 20

4.3

Effect of soil type ......................................................................................................... 20

4.4

Effect of storage temperature ....................................................................................... 21

4.5

Effect of degree of compaction .................................................................................... 22

4.6

Effect of consistent quality .......................................................................................... 22

Results

........................................................................................................................... 23

5.1

Sample preparation ...................................................................................................... 23

5.2

Soil density and strength .............................................................................................. 23
5.2.1 Density of stabiilized mud ................................................................................. 23
5.2.2 Strength of stabilized mud ................................................................................. 23
5.2.3 Density of stabilized peat .................................................................................. 27
5.2.4 Strength of stabilized peat ................................................................................. 30

Swedish Deep Stabilization Research Centre, Report 3

5.3

Comparison with earlier results ................................................................................... 35
5.3.1 Kirkkonummi..................................................................................................... 35
5.3.2 Örebro – Arboga ................................................................................................ 36

Conclusions and recommendations

............................................................................ 39

6.1

Conclusions .................................................................................................................. 39
Density ......................................................................................................................... 39
Shear strength .............................................................................................................. 39
Stabiilizers.................................................................................................................... 39
Storage temperature ..................................................................................................... 40

6.2

Recommendations ........................................................................................................ 40
Stabilizers ..................................................................................................................... 40
Stabilizer quantities ...................................................................................................... 40
Proposed projects ......................................................................................................... 41

Appendix 1

........................................................................................................................... 45

Appendix 2

........................................................................................................................... 46

Appendix 3

........................................................................................................................... 48

Appendix 4

........................................................................................................................... 49

Appendix 5

........................................................................................................................... 50

Appendix 6

........................................................................................................................... 51

Stabilization of Organic Soils

Summary

The purpose of this report is to point out the possibility of stabilizing gyttja and peat by determining
the shear strength of samples stabilized in the laboratory. The report has also served as a basis for
an in depth study at the Swedish Deep Stabilization Research Centre.

Samples of stabilized soil were produced in the laboratory. The soil consisted of gyttja and peat
from two sites and the strength of the stabilized samples was determined with unconfined
compression tests. Different binders were used, consisting of cement (four different types), lime
and residual material from industry.

The main conclusions are as follows:

• Strength both in gyttja and peat can be considerably improved by the use of stabilizers.

• Density tends to increase in the stabilized samples. Increased density also tends to give increased

strength.

• The storage temperature is of varying importance for different stabilizing agents. To make the

best comparison between laboratory stabilization and in situ stabilization, the laboratory samples
must be stored at the same temperature as is expected in the soil in situ.

The results from the stabilization tests show that different types of cement and different mixes with
cement and other stabilization agents vary with respect to effect. Swedish rapid hardening cement
gave the best results in gyttja, but in some samples Swedish standard cement gave equivalent results
at 26 days. Cement mixed with ground granulated blast furnace slag gave gave the next best results
after cement alone.

Cement mixed with ground granulated blast furnace slag gave the best results in peat. Stabilization
with cement alone also gave a good result. Lime has a poor effect in peat.

In almost every case, mixes with fly ash gave the worst results of all residual materials.

The results correspond well with those that can be expected from the theoretical evaluation in
Chapter 4.

Based on the theoretical valuation and the results in the report, Chapter 6.2 recommends that the
binding effect of the stabilizing agent should be based on cement and puzzolanic reactions. With
respect to technical demands, a preliminary investigation is always made. Today, it is proposed that
Portland cement, lime and ground granulated blast furnace slag be approved. Suitable amounts of
binder are suggested.

Swedish Deep Stabilization Research Centre, Report 3

The characteristics and composition of the residual material vary partly with the type of raw
material and the fact that various industrial processes are not designed to produce the best residual
material. Other materials, including new types, should therefore always be reviewed for each
project in accordance with environmental regulations.

A number of projects for further research are presented below:

• Establish and clarify performance and methodology when samples with stabilized soil are being

prepared and tested.

• Examine the importance of the robustness of the stabilizing agents (Chapter 4.6).

• Develop methods and criteria for evaluating the environmental effects of new stabilizing agents.

• Study the chemical and physical parameters affecting reactions that increase strength.

Stabilization of Organic Soils

1. Introduction

1.1

PURPOSE

The purpose of this feasibility study was to demonstrate the feasibility of stabilizing mud and peat
by determining the shear strength of laboratory-stabilized samples. A second purpose of the study is
to serve as a guide to methodology and trial planning for the further development of mass stabiliza-
tion and for a further study planned at the Swedish Deep Stabilization Centre (SD).

1.2

BACKGROUND

In late 1995 the Swedish Geotechnical Institute (SGI) published its report on a major project on the
effects of cement and lime in the deep stabilization of different soils [5]. The project yielded results
that should permit the stabilization of organic soils such as mud and peat. At the same time, mass
stabilization has been introduced in Sweden.

Mass stabilization is a new technique for the stabilization of loose soil layers such as peat and mud.
In mass stabilization, unlike the deep mixing method, binder is blended into the whole layer, result-
ing in a stabilized “block”.

The mass stabilization method has been developed over the last five years and applied with good
results in a number of projects in Sweden and Finland [10, 15]. In Finnish projects using mass sta-
bilization, shear strength has been increased by factors of up to 40 in mud and up to 20 in peat.

SD is carrying on R&D efforts for the further development of deep stabilization technology using
columns of lime cement. In view of the growing interest in the new mass stabilization technique
and its high development potential, it was decided to draw up a separate R&D plan for this tech-
nique. The plan [12] mentions the need to demonstrate and further investigate the effectiveness of
stabilization of organic soils such as mud and peat.

The work presented herein was carried out in 1996 and the report reflects the state of our knowl-
edge at that time. Since then the findings have been actively applied at SD and in the EU project
“EuroSoilStab”.

Swedish Deep Stabilization Research Centre, Report 3

Fig. 1. Mass stabilization.

Spreading of geotextile
and 0.5 m sand on
mass stabilized surface

Gyttja

Clay

Unstabilized soil

Mixing

Mixing tool

Finished
embankment

When a road or railway embankment is to be constructed on clay soil, some form of soil improve-
ment is usually needed to avoid settlement and stability problems. The traditional solution is deep
stabilization of the clay. However, the clay is often overlaid by loose layers of peat and mud in
which it is difficult to achieve adequate bearing capacity by deep mixing. A conventional solution
here is e.g. soil substitution, which involves excavating the loose soil layers and replacing them
with frictional material of higher bearing capacity.

Soil substitution is expensive and frequently also problematic, as the replaced material must be dis-
posed of and new filling material hauled to the site. There has therefore been a need to develop a
functional, economical, and more environmentally friendly method for stabilizing mud and peat.
The recently developed mass stabilization technique meets these requirements.

Mass stabilization is a soil reinforcement technique in which stabilizing agents are blended into the
entire soil layer. Unlike the deep mixing method, this results in a stabilized “block”. Stabilization
reduces subsidence and improves the stability of the soil. Mass stabilization enables loose soil lay-
ers to be used instead of being disposed of. It also reduces extraction and haulage of natural gravel
and other fill.

Hitherto, mixing has been done by means of a tool mounted on an excavator. A geotextile is then
spread over the stabilized surface, followed by a 0.5 m layer of gravel e gravel bed compacts the
stabilized material and is also used to form a working surface for the excavator.

2. Mass stabilization

Stabilization of Organic Soils

As at 1997 the method had been used in more than ten projects in Sweden and Finland, primarily in
road and railway embankments. The binders used in these projects consisted of cement and mix-
tures of cement and various types of industrial residues. Inert soils such as sand can also be mixed
with cement, and various salts can be added to control the stabilization reactions.

Other possible applications for the mass stabilization technique include stabilization of intractable
excavated material to render it usable as fill or to produce an artificial “crust”.

Swedish Deep Stabilization Research Centre, Report 3

The present feasibility study began in April 1996 with mixing trials in the SGI Laboratory at
Linköping. Soil samples of mud and peat were taken from two locations and a number of binders
used for mass stabilization.

Distinct mixing and storage procedures were used for the mud and peat samples. The mud was
mixed and stored in the conventional way in sample tubes of inside diameter 50 mm, while the peat
samples were stored in sample tubes of inside diameter 68 mm and with a vertical load. The mixing
and storage procedure for the peat samples conformed to procedures developed in Finland, which
were demonstrated by Elina Parkkinen of Lahja Rudus Oy Ab. The new test equipment was fabri-
cated in the SGI workshop on the same principles as the Finnish equipment.

After storage for periods of 14 and 26 days (mud) and 28 days (peat), SGI determined the density
of the samples and carried out unconfined compression tests which were evaluated by SGI Labora-
tory for shear strength. Altogether 84 mud samples and 68 peat samples were studied.

3.1

CLASSIFICATION AND PROPERTIES OF MUD AND PEAT

The properties of the unstabilized soils are shown in Table 1. The samples represent mud and peat
from the third runway at Arland airport and mud and peat from the site of the Örebro–Arboga mo-
torway.

The most prominent difference between the peats is their water content. That of the Arlanda peat is
442 % and of the Örebro peat 1308 % and 1413 %. Since the water content of peat varies widely it
is difficult to state a “normal” value. However, the water content of the Arlanda peat can be consid-
ered unusually low.

Besides the differences in water content the peats also differ notably in their content of organic ma-
terial. The Arlanda peat is lower in organic material and more highly humified than the Örebro
peat.

As well as the characteristics reported in Table 1, the chemical composition of the water phase of
the original soils was determined. This was done by expression of the pore water followed by
chemical analysis.

3. Stabilization trials in mud and peat – execution

Stabilization of Organic Soils

Table 1. Properties of unstabilized soil. Density of peat was assumed to be

1.0 ton/m

and of gyttja 1.2 ton/m

According to 1981 system*

Org.

Water

Cone liquid

Humi-

Soil

cont

content

limit

fication

classifiction

w %

von Post

Örebro

Grey-green clayey gyttja with plant parts

8,0

151

140

le Gy vx

Brown-black low-humified peat

1308

H2–H3

Brown-black peat

1413

H2–H6

Arlanda

Grey-green clayey gyttja with plant parts

205

le Gy vx

Black highly humified peat with

442

259

plant parts and wood remnants

1) Determined by the colorimeter method
2) Determined by the ignition method

Based on visual soil classification adjusted for existing measurement data.

3.2

PORE WATER ANALYSIS

Before deep stabilisation is applied in a project, routine laboratory tests are done using various sta-
bilizing agents in order to predict the effectiveness of stabilization. The geotechnical properties of
the unstabilized soil are also determined, including its water content, density, liquid limit, sensitivi-
ty, shear strength and soil type. In some cases its organic content and degree of humification are
also studied.

Studies show that soils with apparently similar mechanical properties can give widely differing
effectiveness of stabilization with the same agents. This is probably due to the effect of different
chemical substances in the soils. Chemical analyses of the unstabilized soil can therefore provide
useful additional information.

One method for the chemical analysis of soil is pore water expression, in which the ion concentra-
tions in the water phase of a stabilized or unstabilized soil are determined. In this feasibility study
pore water expression and analysis were done on the unstabilized soil. The results are shown in
Appendix 5.

The chemical laboratory analysis of stabilized and unstabilized soils is described in more detail in
/ref 5/. The report describes, among other things, the changes in the ionic composition of the water
phase on addition of various stabilizing agents and their effects on strength-enhancing reactions.

Swedish Deep Stabilization Research Centre, Report 3

Fig. 2. Test programme.

Strength
Density

Cement 1 - 4

Lime

Blast furnace slag 1 & 2

Fly ash 1-2

By-pass ash

Fine sand

Örebro

Gyttja
Peat 1 & 2

Arlanda

Gyttja
Peat

Table 2. Stabilizers used.

BINDERS

ADDITIVE MATERIALS

Puzzolanic materials

Latent hydraulic materials additive

Fillers

Chemical

CEMENTS

By-pass ash (B)

GROUND GRANULATED

Fine sand (FS)

Glorit (G)

Cem SH P (SH)

ASH

BLAST FURNACE SLAG

Cem Std P (Std)

Fly ash 1 (F1)

Granulated blast furnace slag 1 (M1)

Pikasementti (FSH)*

Fly ash 2 (F2)

Granulated blast fuurnace slag 2 (M2)

Cement 4 (C4)

LIME

Quicklime (CaO)

* Finnish rapid hardening cement

3.3

STABILIZING AGENTS

The stabilizing agents (Table 2) comprised four types of cement, lime, ground granulated blast fur-
nace slag, fine sand, and various industrial residues. See further Chapter 4.1. In addition, a chemical
addition, Glorit, was tried. All the materials were from sources in the Nordic countries, but not all
of them are commercially available.

The choice of stabilizing agents was based on the results of previous trials in Sweden and Finland.
Altogether 11 stabilizing agents and one chemical addition were used in 19 combinations (see Ap-
pendix 1).

3.4

TEST PROGRAMME

The combinations of binding agents and soils that were tested are shown in summary in Fig. 2. The
test programme in its entirety will be found in Appendix 2.

Stabilization of Organic Soils

The quantity of stabilizing agent used in all the mud samples was 200 kg/m

. The quantity of stabi-

lizing agent added to the peat samples was in most cases 250 kg/m

, but this was varied in some

samples as may be seen from the test programme in Appendix 2.1.

The Arlanda peat was found to be very dry during admixture of stabilizing agent, which made the
stabilized mass crumbly and difficult to mix. Water was therefore added to some of the samples in
order to enhance mixing and to observe the effect on stabilization effectiveness.

3.5

SAMPLE PREPARATION AND TESTING

To reduce the differences in soil properties between samples, each soil was blended until the mass
was relatively homogeneous. After homogenization, which was done by the SGI Laboratory, addi-
tion of stabilizers took place.

Mixing and storage of the mud samples was carried out by the normal procedure for mixing and
storage of clay samples. Under this procedure, the soil and the stabilizing agent were mixed to a
homogeneous mass by means of a household mixer and then compacted and stored in tubes of in-
side diameter 50 mm. In the case of the Örebro mud each layer was compacted by hand, which is
the compacting procedure normally used in Finland. The Arlanda mud was compacted with a con-
stant pressure for approximately 10 seconds after each layer, this being the usual method in Swe-
den. The test results were evaluated by the same method used in unconfined compression testing of
stabilized clay samples. The shear strength is calculated as half of the compressive strength, cor-
rected for deformation according to Swedish standard.

The samples were stored at 8 ºC, after which the shear strength was determined by unconfined com-
pression testing. To provide an idea of the change in the shear strength of the mud over time, com-
pression tests were carried out after 14 and 26 days. The test at 14 days was done on one sample
and the test at 26 days on dual samples.

Mixing and preparation of the peat samples was done by a different method than the mud samples.
The tubes used had an inside diameter of 68 mm and the samples were subjected to a vertical load
during storage. The procedure is described in Fig. 3.

1. A net was taped over the bottom of the tube to permit the peat specimen to take up water during

storage.

2. The peat was mixed with stabilizer to form a “homogeneous” mass and then compacted into the

sample tubes by hand.

3. The specimen was subjected to a dispersed pressure of ~18 kPa by means of an iron cylinder.

4. The stabilized specimens were then placed in a specially design sample box consisting of a plas-

tic tray positioned so that the specimens could be stored vertical. The bottom of the tray was
then filled with water to a depth of ~50 mm.

Swedish Deep Stabilization Research Centre, Report 3

5. After 28 days at room temperature the specimens were expelled from the tubes and unconfined

compression tests performed.

The peat specimens were stored for 28 days and the compression tests were done on dual speci-
mens. Most of the specimens were stored at room temperature (~21 ºC), as has been the practice in
Finland, but because the temperature in the ground is normally ~8 ºC a number of parallel peat
specimens were prepared and stored at 8 ºC.

This method of mixing and storing peat specimens, developed in Finland, is the one that is now
used for testing. The test method is intended to resemble the field conditions of mass stabilization.
The larger diameter of the specimens is intended to provide a more representative sample of the
heterogeneous peat material. The loading imposed after mixing is intended to correspond to a field
loading ~1 m of frictional material. The stabilized peat is enabled to absorb water through the net at
the bottom of the tube. This resembles the field conditions of the stabilized peat at the edge of the
stabilized volume, which can absorb water from adjacent unstabilized peat.

Figure 3. Mixing and storage ppproceduure in peat tests.

Peat mixed
with stabilizer

Iron cylinder,
m = 6.5 kg

(18 kPa)

d : H

≈≈≈≈

1 : 2

~50 mm
water

Stabilization of Organic Soils

Numerous factors affect the strength obtained on stabilization. The variation in stabilization effec-
tiveness between different types and quantities of stabilizer is normally greater in stabilized muds
and peats than in stabilized clays.

This chapter describes briefly how various factors in principal can influence stabilization effectiveness.

4.1

EFFECT OF STABILIZERS

When stabilizers are added to a soil layer, various reactions take place whereby the soil is bound
together and its strength is increased. Different stabilizers build up strength in different ways. The
main admixtures used for the deep mix stabilization of clays in Sweden are lime and cement. How-
ever, other binders/fillers may be appropriate, just as in other types of soil stabilization. The stabi-
lizers discussed in connection with mass stabilization are divided into the following groups:

• binders: cement and quicklime
• latent hydraulic admixtures, e.g. ground granulated blast furnace slag
• pozzolanic admixtures, e.g. fly ash
• fillers, e.g. fine sand

The group to which a stabilizer belongs depends, in simple terms, on its CaO:SiO

ratio, its general

mineralogical composition, and its particle size and shape (see Table 3).

4. The influence of various factors

on stabilization effectiveness

Table 3 Properties affecting reactivity of stabilizers (after /ref 4/).

Classification

Chemical compo-

Mineraological

Particle

sition CaO/SiO

structure

Binders
Portland cement

~ 3

crystalline

~ 300 – 500 m

/kg

Quicklime

> 40

0 – 0.1 mm

Latent hydraulic additives
Ground granulated

~ 1

amorphous

~ 400 – 600 m

/kg

blast fuuuurnace slag

Puzzolanic additives
Coal fly ash

~ 0.1 – 0.5

amorphous/crystalline

~ 300 – 500 m

/kg

Fillers
Fine sand

<< 0.1

kristallin

0.006 – 0.002 mm

Swedish Deep Stabilization Research Centre, Report 3

4.1.1

Cement and lime

Cement and lime are binding agents. They bind and strengthen the soil without the need to add an
activator. The strength enhancing reactions are briefly discussed below:

• Lime is produced by burning limestone. It reacts immediately on contact with water in the soil to
form slaked lime or calcium hydroxide (Ca(OH)

). This reaction generates heat and the pH value

increases to approximately 12.5. It is a condition for the subsequent pozzolanic reactions, in which
clay particles in the soil react with the calcium hydroxide forming strength enhancing reaction
products.

• Cement consists of numerous minerals and is manufactured by combining cement clinker (a sin-
tered material of limestone and clay) with gypsum. Cement mixed with water forms calcium silicate
hydrate and calcium hydroxide (Ca(OH)

). Calcium silicate hydrate, generally referred to as CSH

gel, forms on the surfaces of the cement particles and because it has a strongly cementing effect it
binds the soil together and increases its strength. Since the hydraulic reaction takes place considera-
bly faster than the pozzolanic reaction, cement stabilized soil normally attains higher strength than
lime stabilized soil, particularly in the first few months.

Since some Ca(OH)

is formed during cement stabilization, pozzolanic reactions will also take

place, though to a lesser extent than in lime stabilization. Hence in cement stabilization, in addition
to the cementation reaction, the same strength enhancing reaction products are formed as in lime
stabilization in about one fifth of the quantity.

• In conventional deep stabilization by deep mixing, the binder combination cement and lime has
been bound to give good stabilization effectiveness. Cement provides rapid, high and “robust”
growth in strength. The slaking of the lime also provides momentary drying and heat evolution
which accelerates the reaction of the cement, while the pozzolanic reactions provides a further in-
crement of strength in the longer term.

Examples of the chemical composition of cement and lime will be found in Table 4.

Table 4. Examples of percentage chemical composition of standard Portland cement, lime,

and granulated blast furnace slag.

Portland cement

Quicklime

Granulated

(Slite Std)

blast furnace slag

SiO

0.8

—

CaO

MgO

1.6

—

Loss on ignition

1.6

—

Stabilization of Organic Soils

4.1.2

Fly ash

Fly ash is obtained as a residue from power stations and heating plants fuelled with pulverized fuel.
The properties and reactivity of the ash vary depending on e.g. the coal and the combustion process.

Almost all the fly ash obtainable and used in Nordic countries comes from coal combustion. This
ash is low in CaO and gives pozzolanic reactions when used as a stabilizer. Pozzolanic stabilizers
do not react by themselves during stabilization, but they can form strength enhancing materials very
slowly on addition of water and some form of calcium hydroxide (Ca(OH)

). Calcium hydroxide is

normally added in the form of Portland cement in order to obtain faster “normal” strength enhance-
ment.

Since fly ash is a residue from a process whose primary purpose is to produce energy, even ash
from a single plant can have widely varying properties. A plant will not always use the same grade
of coal, or even coal from the same deposit, which naturally results in wide variations in the techni-
cal and environmental properties of the residue. Each batch must therefore undergo technical and
environmental quality assessment.

4.1.3

Granulated blast furnace slag

During the refining of ore into metal various types of slag are formed as by-products. Granulated
blast furnace slag, a by-product of iron smelting, is the slag most frequently used for mass stabiliza-
tion and is therefore the main focus of the present discussion. Examples of chemical compositions
of granulated blast furnace slag will be found in Table 4.

The mineralogical composition and reactivity of the slag vary depending on the rate of cooling after
leaving the blast furnace (1500 ºC). Slow cooling results in a crystalline, quite unreactive slag,
while rapid cooling produces an amorphous, glassy, latent reactive slag. Granulated blast furnace
slag is a product of fast cooling and is thus a latent reactive slag.

Granulated blast furnace slag used as a stabilizer has latent hydraulic properties. This means that,
like pozzolanic materials, the slag can form strength-enhancing products with calcium hydroxide
(Ca(OH)

). The difference is that the slag contains rather more reactive lime. However, the reaction

rate of the slag itself is so slow as to be negligible. Some form of activitation is therefore necessary
if it is to be used as a binder. The most commonly used activator for slag is Portland cement. Other
substances can be used as activators, but their reactions are quite different and the reaction products
have different properties and durability. The use of other activators should therefore be discussed
and clarified in advance.

As with other pozzolanic materials, the strength-enhancing reactions that occur during stabilization
with granulated blast furnace slag are highly temperature-sensitive. Higher temperatures normally
increase the reaction rate and hence the strength. Conversely, the strength enhancement normally
falls drops rapidly if the temperature falls.

Availability of usable slag is usually dependent on proximity to a smelting works. Since granulated
blast furnace slag is merely a by-product, one has to accept the composition supplied. There is nor-

Swedish Deep Stabilization Research Centre, Report 3

mally little variation in quality in slag from the same smelting works if the raw material is constant.
However, product declarations and even quality are an essential condition for the use of granulated
blast furnace slag. Swedish engineers have a long record of good experience with the product Merit
5000, binder M1 in the present report.

4.1.4

Filler materials

To increase the number of solid particles a filler, such as fine sand, may be added in soil stabiliza-
tion. The filler itself does not react but increases the strength of the soil by acting as a “stiffener”.

The filler material will be of greatest relevance in the stabilization of peat and mud, as these soils
often require large quantities of stabilizers (see Chapter 4.2). Replacing part of the stabilizer with
inexpensive filler can save costs. The filler may also be expected to fill any voids formed during
stabilization.

In practice, fillers do differ in effectiveness since no filler is completely inert. Thus, for example,
high-silica sand is likely to have a greater effect than limestone filler. However, the effect of fillers
of whatever type is considerably less than that of the same quantity of binder.

4.2

STABILIZER QUANTITIES

Peat and mud normally require greater quantities of stabilizer than does clay. This is partly because
peat and mud contain fewer solid particles to stabilize. Since it is the solid particles that provide
structure, a greater quantity of stabilizer needs to be added. Moreover, mud and peat have a consid-
erably higher water:soil ratio than clay. The large amount of water in the soil implies larger voids,
requiring more stabilizer.

4.3

EFFECT OF SOIL TYPE

Mud and peat, unlike clay, have high organic content. The organic material may include retarding
substances such as humus and humic acids. During stabilization the humic acids react with
(Ca(OH)

) to form insoluble reaction products which precipitate out on the clay particles. The acids

may also cause the soil pH to drop. This negatively affects the reaction rate of the binders, resulting
in a slower strength gain in mud and peat than in clay.

Studies in Finland /ref 16/ indicate that in soils with high organic contents, such as mud and peat,
the quantity of binder needs to exceed a “threshold”. As long as the quantity of binder is below the
threshold the soil will remain unstabilized. A reason for this may be that the humic acids are neu-
tralized when sufficient binder is added.

A recent study at the University of Oulu, Finland, /ref 7/ shows the negative effect of humus and
humic acids on the effectiveness of soil stabilization. However, the results of the study indicate that
the humus and humic acid content of the soil is only one of several factors affecting stabilization
effectiveness. Hence the stabilization outcome of a binder cannot at present be definitely predicted
merely by determining the organic content and humus content of the soil.

Stabilization of Organic Soils

Cement is often a more effective stabilizer than lime in mud and peat soils. This is probably due to
the effect of humic acids as discussed above and to the inhibition of one of the most important
strength-enhancing mechanism of the lime (pozzolanic reactions). In pozzolanic reactions the lime
reacts with clay particles in the soil to form binding materials. In peat and mud the organic material
occupies so much of the soil volume that the stabilizer fails to come in contact with the few clay
particles that are present, with the result that pozzolanic reactions do not take place. Cement gives a
more robust strength gain as the cement forms binding materials with water and clay particles play
no role.

4.4

EFFECT OF STORAGE TEMPERATURE

The heat evolved when a soil is stabilized depends on the stabilizer used (see Fig. 4). Heat acceler-
ates the reaction rate of stabilization.

The temperature in the soil is normally around 8 ºC. At this temperature most stabilizers react slow-
ly. The strength enhancing reactions that are least affected by ambient temperature are the hydraulic
reactions.

One difference that exists between stabilizers is that those with high heat evolution (e.g. lime and
cement) are less dependent on the temperature of the ambient soil than those having slow and low
heat evolution in themselves (e.g. granulated blast furnace slag). Hence in a soil stabilized with low
exothermic agents the reaction rate, and hence the short-term strength, will fall if the ambient tem-
perature drops. Similarly, the strength will increase if the ambient temperature rises. Hence for best
comparison with real conditions laboratory samples should be stored at temperatures corresponding
to field temperatures, i.e. different storage temperature for different types of stabilizers.

Fig. 4. Heat development in soil after stabilization with variout stabilizers.

Swedish Deep Stabilization Research Centre, Report 3

4.5

EFFECT OF DEGREE OF COMPACTION

The bulk density of mud and peat is normally very low, i.e. the ratio of voids to solids is relatively
high. The voids are mostly filled with water. The density of mud and peat normally tends to in-
crease on stabilization since some of the water in the soil is replaced by the stabilizer. This also
reduces the voids fraction, particularly in the case of peats containing large amounts of water.

Since strength of stabilized material generally increases as the voids fraction decreases (other con-
ditions being equal), the effectiveness of stabilization in peat and mud is likely to depend on how
well compacted the material becomes. Laboratory mixing tests show large differences in stabiliza-
tion effectiveness between peat specimens stored under load and specimens stored without load.
One reason is that peat often gets very sticky during mixing, making it difficult to compact. Storage
under load expels any air pockets and hence higher strength is attained.

In order for the stabilizer to react completely it is also important for the stabilizer to be homogene-
ously mixed with the soil. In general, stabilization effectiveness increases with the homogeneity of
the stabilized material.

4.6

EFFECT OF CONSISTENT QUALITY

To ensure homogeneity of the mass stabilized volume it is important that the stabilization method is
relatively insensitive to variations in e.g. soil and execution. Stabilizers vary in their ability to yield
a product of consistent and reliable quality.

The commonest stabilizing agents today, cement and lime, have been found to yield a product of
consistent and reliable quality in “normal soils”, see e.g. /5/. Given that mass stabilization is often
done in organic soils in which a certain amount of water flow may occur, the consistent quality,
reaction rate and environmental properties of a stabilizer are of still greater importance than in con-
ventional deep stabilization of clay.

Stabilization of Organic Soils

5. Results

This chapter reports the results of the stabilization tests on mud and peat. At the end of the chapter
we compare our results with those of some earlier experiments.

5.1

SAMPLE PREPARATION

With mud the addition of stabilizer and preparation of specimens was relatively straightforward.
There was no noticeable difference in sample preparation between the Arlanda and Örebro muds or
between the different stabilizers.

The peats were more difficult than the muds to mix into a homogeneous mass. While the Örebro
peat, with a water:soil ratio of over 1300 %, was relatively easy to mix and compact, the Arlanda
peat, with a water:soil ratio of ~ 400 %, was crumbly and dry even before stabilization. Adding
stabilizer made the sample still drier, which strongly affected the result. Mixing and compaction
proceeded considerably more smoothly in the Arland peat samples to which water was added.

5.2

SOIL DENSITY AND STRENGTH

Chapter 5.2 reports the density and the results of unconfined compression tests on the mud and
peat. The complete results are presented in Appendices 3 and 4. Strength testing of the mud was
done after storage for 14 and 26 days and of the peat after 28 days. In general, the stabilization ef-
fectiveness achieved was very good.

5.2.1

Density of stabiilized mud

The density of both the Arlanda mud and the Örebro mud increased somewhat after stabilization,
see Figs 5a and 5b. The density increase was slightly greater in the Örebro mud, which was com-
pacted manually, than in the Arlanda mud, which was compacted by constant pressure. There is no
discernible difference in the density increase depending on the stabilizer added.

5.2.2

Strength of stabilized mud

Conventional binders

Jet cement by itself results in the best strength in both Arlanda and Örebro mud, see Figs 6 and 7.

A stabilizer consisting of jet cement mixed with lime gives relatively low shear strength at 14 days.
The shear strength at 26 days is somewhat higher in the Arlanda mud but is still considerably lower
than that obtained by stabilization with jet cement alone, see Fig. 6.

Results similar to those reported above are also reported in /ref 5/. However, /ref 13/ recommends a
mix of 50 % cement and 50 % lime for the deep mix stabilization of mud. This is justified by the

Swedish Deep Stabilization Research Centre, Report 3

a) * Compaction manually

Fig. 5.

Density of stabilized gyttja. The 26 day results are averages of two specimens.
200 kg stabilizer per m

was used in all specimens.

(a) Arlanda (compacted by constant pressure unless otherwise indicated)
(b) Örebro (compacted manually)

Densitet gyttja Arlanda

0,2

0,4

0,6

0,8

1,2

1,4

1,6

0 S

+2%

0 S

+ M

0 C

0 F

+ M

0 S

+ M

0 S

sit

t [

Unstabilized gyttja

Densitet gyttja Örebro

0,2

0,4

0,6

0,8

1,2

1,4

1,6

0/5

0 S

+2%

0 S

td +

0 S

t [

]

Unstabilized gyttja

Density Arlanda gyttja

Density Örebro gyttja

Density [kg/m

]

Density [kg/m

]

Stabilization of Organic Soils

Fig. 6.

Strength evolution on stabilization with rapid hardening cement alone and with
rapid hardening cement and lime. 200 kg stabilizer per m

was used in all speci-

mens.
a) Arlanda gyttja

b) Örebro gyttja

Development of shear strength

– stabilized gyttja from Arlanda

100

200

300

400

500

600

Time [days]

r s

engt

h [k

50/50
SH+CaO

Development of shear strength

– stabilized gyttja from Örebro

100

200

300

400

500

600

Tim e [days]

ear

[

50/50
SH+CaO

extensive fund of experience of deep mix stabilization and the robustness of the mixture.

Other stabilizers

By comparing samples stabilized with jet cement alone (reference samples) and mixtures of jet ce-
ment with other materials the relative effects of different stabilizers can be determined. Such a com-
parison is shown is Fig. 7.

The best stabilization effectiveness in both Arlanda and Örebro is obtained with jet cement alone.
The results with mixtures are inferior to those with jet cement alone in every case. Of industrial
residues, best results are obtained with ground granulated blast furnace slag and bypass ash.

Comparison of different cements

Different cements can be compared with each other in the same way as the industrial residues.
Fig. 8 shows a comparison of stabilization tests using 1:1 mixtures of different cements with ground
granulated blast furnace slag.

The results indicate that stabilization effectiveness varies with the cement. The cement that gave
best and most reliable results with both the Arlanda and the Örebro mud was Swedish jet cement.
With Arlanda mud, Swedish jet cement and standard cement and Swedish jet cement gave equal
stabilization effectiveness at 26 days, while cement 4 gave poorer results. The Örebro mud is most
effectively stabilized by Swedish jet cement, Fig. 8.

Swedish Deep Stabilization Research Centre, Report 3

Stabilized Örebro gyttja

100

200

300

400

500

600

(

60/4

0 S

50/5

0 S

/50

+ B

0 S

r s

tren

[

14 dygn

26 dygn

a) * Compacted manually

b) (1) Wide dispersion between specimens.

Stabilized Arlanda gyttja

100

200

300

400

500

600

50/

60/

40 S

50/

50 S

/40 S

/50 S

60/

40 S

50/

50 S

50/

50 S

50/

50 S

ear

[

14 dygn
26 dygn

Fig. 7.

Shear strength of gyttja after storage for 14 and 26 days. The 26 day results are
averages of two specimens. 200 kg stabilizer per m

was used in all specimens.

a) Arlanda (compacted by constant pressure unless otherwise indicated)
b) Örebro (compacted manually)

days
days

days

Stabilization of Organic Soils

Fig. 8.

Shear strength of gyttja after storage for 14 and 26 days. The 26 day results are
averages of two specimens. 200 kg stabilizer per m

was used in all specimens

a) Arlanda (compacted by constant pressure)
b) Örebro (compacted manually)

Stabilized Arlanda gyttja

100

200

300

400

500

600

50/50

SH+M1

50/50

Std+M1

50/50 C4 +

50/50 FSH

+ M1

ear

[

14 dygn

26 dygn

Stabilized Örebro gyttja

100

200

300

400

500

600

/50 S

50/

td+

0 C

4 +

)

r s

tre

[

kPa

]

14 dygn

26 dygn

b) (1) Wide dispersion between specimens.

Glorit chemical admixture

A number of parallel specimens were prepared with a chemical admixture of 2% Glorit (the trade
name of an inorganic salt), Fig. 9. The results show no absolute difference between specimens with
and without Glorit in many cases. In other cases strength values with Glorit are slightly better or
worse. Overall, based on these results, Glorit seems to have no effect.

5.2.3

Density of stabilized peat

The stabilized Örebro peat specimens showed a uniform density increase regardless of what stabi-
lizer was added, Fig. 10b. The density of the Arlanda peat specimens, on the other hand, tended to
decrease or remain unchanged after stabilization, Fig. 10a. Exceptions were the Arlanda peat speci-
mens to which water was added: the density of these specimens increased.

days

Swedish Deep Stabilization Research Centre, Report 3

Fig. 9.

Effect of Glorit (G) on stabilization effectiveness in gyttja. The 26 day results are
averages of two specimens. 200 kg stabilizer per m

was used.

a) Arlanda

b) Örebro

b) (1) Wide dispersion between specimens.

Stabilized Arlanda gyttja

100

200

300

400

500

600

+ 2

0 S

/50 S

50/

50 S

0 S

+ 2

ear

[

14 dygn

26 dygn

Stabilized Örebro gyttja

100

200

300

400

500

600

50/

/50 S

0 S

2 +

0 S

/50

)

r s

tre

[

14 dygn

26 dygn

days

days
days

Stabilization of Organic Soils

Densitet torv Arlanda

0,2

0,4

0,6

0,8

1,2

1,4

(

5 k

3 (

(

)

0 k

3 (

(

)

td +

/50 F

ens

t [

]

Unstabilized peat

b) (*) Stored at 8 ºC (other specimens were stored at 21 ºC).

Fig. 10. Density before and after stabilization. 250 kg stabilizer per m

was used unless

otherwise indicated. Results normally represent the average of two specimens.
a) Arlanda peat

b) Örebro peat 1 and 2

a) (*) Stored at 8 ºC (other specimens were stored at 21 ºC) (**) Water added

De nsite t torv 1 Öre bro

0,2

0,4

0,6

0,8

1,2

1,4

g/m

50/

0 S

t [

]

Ostabiliserad torv

Densitet torv 2 Örebro

0,2

0,4

0,6

0,8

1,2

1,4

(

50/

50 S

60/

40 S

50/

50 S

50/

50 S

1 (

50/

50 S

50/

50 S

/50 S

td +

50/

4 +

50/

50 F

50/

50 S

50/

[

Ostabiliserad torv

Density Arlanda peat

Density peat 2 Örebro

Density Örebro peat 1

Unstabilized peat

Density [kg/m

]

Density [kg/m

]

Density [kg/m

]

Swedish Deep Stabilization Research Centre, Report 3

5.2.4

Strength of stabilized peat

The Arlanda peat was “crumbly” in the unconfined compression tests. The results show very low
shear strength regardless of what stabilizer was added. The hypothesis was formed that too little
water was present for all the binder to react. The specimens to which water was added showed good
stabilizing effectiveness, Fig. 11, confirming the presence of unreacted binder (cement) in the spec-
imens. The water addition enabled hydraulic reactions to take place.

Conventional binders

In the Örebro peat high shear strength is obtained with most stabilizers. However, a mixture of jet
cement with lime gives relatively low strength, Fig. 12.

Other stabilizers

Fig. 13 compares the relative effectiveness of different stabilizers. Samples stabilized with jet ce-
ment alone (reference samples) are compared with samples stabilized with 1:1 mixtures of jet ce-
ment with other materials.

The best stabilizing effectiveness in the Örebro peat is obtained with a mixture of jet cement and
granulated blast furnace slag. Stabilization with jet cement alone is also very effective. Other mix-
tures yield poorer results.

Storage temperatures

The specimens that were stored at 8 ºC instead of 21 ºC confirm that the storage temperature plays a
greater role in stabilization with granulated blast furnace slag than with jet cement alone. After stor-
age at 8 ºC, better stabilization effectiveness was obtained with jet cement alone than with mixtures
of jet cement and granulated blast furnace slag. In specimens stored at 21 ºC, somewhat better stabi-
lization effectiveness is obtained with mixtures of jet cement and granulated blast furnace slag,
Fig. 14.

Comparison of different cements

Fig. 15 compares different cements. The specimens were stabilized with 50:50 mixtures of various
cements with granulated blast furnace slag No. 1.

The results show that the stabilization effectiveness varies with the cement. With the Arlanda peat,
jet cement was somewhat more effective than other cements. With the Örebro peat standard cement
gives somewhat better stabilization effectiveness than jet cement. Swedish standard cement and jet
cement are both more effective than Finnish jet cement and cement 4.

Glorit chemical admixture

To investigate whether the chemical admixture of Glorit affects stabilization effectiveness with
peat, parallel speciments were prepared with jet cement alone and jet cement plus 2 % Glorit. On
the basis of the results in Fig. 16, Glorit does not appear to have any positive effect.

Stabilization of Organic Soils

Fig. 12. Shear strength on stabilization with jet cement alone and with rapid hardening

cement and lime. 250 kg stabilizer per m

was used unless otherwise indicated.

Results normally represent the average of two specimens
a) Arlanda peat b) Örebro peat 2

(*) Stored at 8 ºC (other specimens were stored at 21 ºC) (**) Water added

Fig. 11. Effect of water addition to Arlanda peat. 250 kg stabilizer per m

was used unless

otherwise indicated. Results normally represent the average of two specimens.

Stabilized Arlanda peat

100

200

300

400

500

5 k

g/m

3 (

125 k

3 (

(

**)

250 k

3 (

(

**)

28 day shear

ngt

h [

Pa]

Stabilized Arlanda peat

100

200

300

400

500

50/

ay sh

r st

[

Stabilized Örebro peat 2

100

200

300

400

500

50/

50 S

28 d

ay sh

ear

[kP

Swedish Deep Stabilization Research Centre, Report 3

Fig. 13. Shear strength after storage at 21 ºC for 28 days. 250 kg stabilizer per m

was used

unless otherwise indicated. Results normally represent the average of two speci-
mens
a) Arlanda peat

b) Örebro peats 1 and 2

Stabilized Arlanda peat

100

200

300

400

500

0 S

50 SH

50/

50 SH

r st

[

Stabilized Örebro peat 2

100

200

300

400

500

50/

60/

40 S

50/

50 S

0 S

50/

50 S

50/

28 day s

r st

[

Stabilized Örebro peat 1

100

200

300

400

500

50 S

50/

0 S

ay sh

r st

[

Stabilization of Organic Soils

Fig. 14. Comparison of stabilized Örebro peat stored at 8 ºC and at 21 ºC.

250 kg stabilizer per m

was used unless otherwise indicated.

Fig. 15. Shear strength after storage for 28 days. 250 kg stabilizer per m

was used unless

otherwise indicated. Results normally represent the average of two specimens.
a) Arlanda peat

b) Örebro peat 2

Stabiliserad torv (2) Örebro

100

200

300

400

500

28 dy

gns skjuvhållf

thet

Pa]

50/50 SH
+M1

Lagring vid 21 grader

Lagring vid 8 grader

Stabiliserad torv Örebro

100

200

300

400

500

50/50

SH+M1

50/50

Std+M1

50/50

C4+M1

50/50

FSH+M1

28 dygns skjuvhållfasthet [kPa]

Stabilized Arlanda peat

100

200

300

400

500

50/50 SH+M1

50/50 Std + M1

50/50 C4 + M1

50/50 FSH + M1

28 day shear strength [kPa]

Stabilized Örebro peat

Stabilized Arlanda peat

28 day shear strength [kPa]

Stabilized Örebro peat 2

28 day shear strength [kPa]

Swedish Deep Stabilization Research Centre, Report 3

Different stabilizer quantities

Specimens of the Örebro peat were prepared with different quantities of stabilizer (jet cement
alone), Fig. 17. The 28 day shear strength of these speciments shows an increase over the range of
addition rates from 70 kg/m

to 250 kg/m

. However, an addition rate of 400 kg/m

gives lower

strength than 250 kg/m

, indicating that the water:binder ratio has an effect on the shear strength.

Figur 16.Effect of Glorit on stabilization effectiveness in peat. 250 kg stabilizer per m

was

used. Results normally represent the average of two specimens.
a) Örebro peat 1

b) Örebro peat 2

Stabilized Örebro peat 1

100

200

300

400

500

r st

[

Stabilized Örebro peat 2

100

200

300

400

500

r s

[

kPa]

Stabilization of Organic Soils

Fig. 17. Shear strength after storage for 28 days. Tests using 250 kg stabilizer per m

were

done on two parallel specimens, the results above representing the average. Other
tests were done on single specimens.

Stabilized Örebro peat 1

100

200

300

400

500

0 k

g/m

50 k

00 k

ay sh

ear

[

5.3

COMPARISON WITH EARLIER RESULTS

5.3.1

Kirkkonummi

Ahead of a road construction project at Kirkkonummi, west of Helsinki, stabilization trials were
done on clay, mud and peat from the site. Selected results of unconfined compression tests on the
mud and peat are reported in Appendix 6 and Fig. 18.

The best stabilization effectiveness was obtained with jet cement mixed with 100 kg/m

of fine sand

and with jet cement alone. The addition of filler (100 kg/m

fine sand) to the jet cement gives an

increment of ~30 kPa in shear strength, Fig. 18. This confirms that the number of solid particles in
the peat has an influence on the effectiveness of stabilization, since the shear strength increases
when fine sand (i.e. solid particles) is added. Of other stabilizers tested, a 50:50 mixture of Finnish
standard cement and granulated blast furnace slag also gives good stabilization effectiveness,
Fig. 18.

Swedish Deep Stabilization Research Centre, Report 3

F* = Finnstabi, special binder from Kemira pigments.

Fig. 18. Shear strength with different stabilizer quantities. Results are from tests ahead of a

stabilization project at Kirkkonummi, Finland. From Finsementti.

Stabiliserad gyttja/torv från Kyrkslätt

100

200

300

400

500

600

700

100

150

200

250

300

350

400

M ängd binde m e de l [k g/m 3]

skj

llf

Finnish SH cem

Finnish std cem

Finnish std cem +
granulated blast
furnace slag 3:7
Finnish std cem +
granulated blast
furnace slag 1:1

SH+100kg/m

fine sand
SH+CaO 3:1
SH+CaO 1:1
SH+CaO+F* 1:1:1

5.3.2

Örebro – Arboga

The feasibility study used mud and peat from a motorway construction site between Örebro and
Arboga. Ground improvement in the form of deep and mass stabilization is planned in a number of
clayey and peaty areas, and stabilization tests were therefore carried out. Selected results are report-
ed in Appendix 6. Fig. 19 compares the results from this feasibility study (peats 1 and 2) with those
of the preliminary investigation (peats A and B).

The results differ in absolute level, which is certainly to be explained by differences in sampling
methods and probably also by differences in the original peats. However, the results do show that in
both cases a mixture of cement and lime gives the poorest outcome, while cement mixed with
ground granulated blast furnace slag and cement alone give the best outcome.

Binder content [kg/m

]

30 day shear strength [kPa]

Stabilized gyttja/peat from Kirkkonummi

Stabilization of Organic Soils

Örebro Peat 2

100

200

300

400

500

600

700

50 100 150 200 250 300

Binder content [kg/m

]

28 day shear

str

ength [kP

50/50 SH+M1

50/50 SH+FS

50/50 SH+F2

50/50 SH+CaO

Örebro Peat 1

100

200

300

400

500

600

700

50 100 150 200 250 300

Binder content [kg/m

]

28 day shear

str

enth [kP

50/50 SH+M1

50/50 SH+FS

50/50 SH+F2

Fig. 19. Results from earlier stabilization tests at Örebro compared with results from this

feasibility study. Shear strength determined by unconfined compression test after
storage for one month.
a) Stabilized peats 1 and 2 from this feasibility study. 250 kg stabilizer per m

was

used.

a) Peats from feasibility study
Peat 1 = Water content: 1308 %

Organic content: 99 %

Peat 2 = Water content: 1413 %

Organic content: 97 %

Swedish Deep Stabilization Research Centre, Report 3

Stabiliserad torv från Örebro-Arboga

100

200

300

400

500

600

700

100

150

200

250

300

350

400

450

Mängd bindem edel [kg/m 3]

juv

hå

llf

]

Std

Torv A

Std+granulerad
masugnsslagg 1:1

SH+CaO 1:1
SH+finnstabi 1:1
Lohjamix V15

Stabiliserad torv från Örebro-Arboga

100

200

300

400

500

600

700

100

150

200

250

300

350

400

450

Mängd bindem edel [kg/m 3]

juv

hå

llf

]

Std

Torv B

Std+granulerad
masugnsslagg 1:1

SH+CaO 1:1
SH+finnstabi 1:1
Lohjamix V15

b) Previously tested peats from Örebro–Arboga:
Peat A = Density 0.98 t/m

Water content: 1350 % Organic content: 99,1 %

Peat B = Density 0.98 t/m

Water content: 1290 % Organic content: 98,9 %

Fig. 19b. Shear strength with different stabilizer quantities. Results are from previous tests

from Orebro – Arboga.

Stabilized peat from Örebro-Arboga

30 day shear strength [kPa]

Peat A

Peat B

Stabilization of Organic Soils

6.1

CONCLUSIONS

The studies show that the strength of peat and mud can be considerably improved by stabilization.
The strength attainable depends on the composition of the added stabilizer and on the choice of
filler, if used.

New method for preparation and storage of stabilized peat specimens
A special method was used for the tests on peat (Chapter 3.6), which was developed and is current-
ly in use in Finland. The special test equipment that it requires was constructed by the SGI work-
shop after studies of the Finnish equipment. Both compaction and storage proceeded relatively
smoothly. Extraction of the speciments from the tubes was considerably more difficult than with
conventional tubes. However, this should be possible to improve.

Density

With the exception of the Arlanda peat, the density tends to increase after stabilization (Figs 6 and
11). A relationship can be seen between the increase in density and an increase in the strength of
the specimens. A possible explanation of this might be that the specimens whose density increased
were well compacted, resulting in reduced pore volume and hence greater strength. No relation can
be seen between the change in density and the stabilizer used.

Shear strength

Unconfined compression tests give very high values of shear strength for a number of stabilizers,
both with Örebro mud and peat and with Arlanda mud. The Arlanda peat, however, gave low shear
strength values. A reason for this may be the low water:soil ratio of the peat and its crumbly con-
sistence. The hypothesis was formed that not enough water was present for all the binder to react.
Good stabilization effectiveness was measured in the specimens to which water was added, con-
firming that unreacted binder (cement) was present in the specimens. Adding water permitted hy-
draulic reactions to take place.

Stabiilizers

The results of the stabilization tests show that different cements and mixtures vary considerably in
effectiveness. In mud, Swedish jet cement alone was gave the most effective stabilization, but in
some cases Swedish standard cement was equally good. All mixtures gave poorer results with mud
than did cement alone. In peat, mixtures of cement and ground granulated blast furnace slag gave
the best results. Stabilization with cement alone (standard cement and jet cement) also gave good
results.

Mixtures containing fly ash gave in almost every case the worst results of all the industrial residues.
In peat, lime gives poor stabilization effectiveness. The results are in good agreement with those

6. Conclusions and recommendations

Swedish Deep Stabilization Research Centre, Report 3

that may be expected on theoretical grounds (see discussion in Chapter 4).

Storage temperature

The role storage temperature varies with different stabilizes. For the best comparison with field
conditions, specimens should be stored at temperatures similar to those that can be expected to oc-
cur in the field.

6.2

RECOMMENDATIONS

The recommendations below are based on currently well-known binder reactions and on current
thinking on traditional deep-mix stabilization.

Stabilizers

Based on the review in Chapter 4 and on the results of this feasibility study, we recommend that
binding effect of stabilization should be based on hydraulic or pozzolanic reactions. A pilot study is
always done in accordance with technical requirements. Materials used today are Portland cement,
lime, inert fillers and ground granulated blast furnace slag. They are provided with product declara-
tions and the environmental characteristics are also well-known. Documentation of these can be
found e.g. in the “building product declarations” (

byggvarudeklaration

) developed by the Swedish

construction industry, see e.g. /ref 17/.

The properties and composition of industrial residues vary, due among other things to different raw
materials and to the fact that industrial processes are not driven by requirements on the composition
of residues. Other materials and new materials should therefore always be reviewed under the
Swedish Environment Act and their reactions, reaction products and their properties should be in-
vestigated in each individual project.

Stabilizer quantities

On the basis of the present study and other experience the authors recommend following binders
and quantities for mud, peat and hydraulic fill:

Soil

Binder

Typical quantity

Gyttja

100 – 200 kg/m

Cement or mixtures of
cement and granulated

Peat

blast furnace slag

150 – 250 kg/m

Hydraulic fill

70 – 200 kg/m

Stabilization of Organic Soils

The above ranges are the normal ones. The final choice of binder and quantity is made on the basis
of a technical pilot study and the conditions of the project.

Proposed projects

A number of projects are proposed below for further work.

• Develop a standardized test method. Today a variety of test methods are in use for peat, result-

ing in wide variations and difficulty of comparing results.

• Investigate the importance “robustness” (Chapter 4.6) of a stabilizer, i.e. its ability to maintain

its reaction rate and provide consistent and reliable quality notwithstanding variations in soil and
work methods. Soil layers containing peat and mud require more “robust” stabilizers than clay
soils due to the presence of humus and the wide variation in their properties.

• Develop methods and criteria for evaluating the environmental effects of new stabilizers.

• Study in detail the chemical and physical parameters that affect strength-enhancing reactions.

Ideally it would be possible to determine the most appropriate stabilizer from a simple chemical
assay of the unstabilized soil.

• Study the effect of natural fillers, such as sand, on the strength of stabilized soils.

• Study the effect of water content on strength. Is there a minimum water:soil ratio below which

stabilization of peat with a dry stabilizer is not possible?

• The mixing, compaction and homogeneity of dry peats may well be improved by adding the

stabilizer as a slurry rather than in dry form. Is it possible to add slurry to dry muds and peats?
Since the water content of peaty soils is subject to seasonal variation, the mixing method (dry or
as slurry) might itself be varied.

Swedish Deep Stabilization Research Centre, Report 3

[1]

Andersson, R., Johansson, S-E., Retelius, A. (1992)

. Rätt valt bindemedel ger bättre

produkt, Cementa AB, Danderyd.

[2]

Andersson, E. (1995)

. Kemisk stabilisering av jord – en studie av samband mellan kemiska/

fysikaliska indexegenskaper och stabiliserinseffekter, Institutionen för teknisk utbildning,
Linköpings tekniska högskola, Linköping.

[3]

Angelva, Huttunen, Kujala (1993)

. Stabilisering av torvmark, Föredragsammanställning,

Uleåborgs Universitet, Uleåborg, Finland.

[4]

Fagerlund, G

. Tidningen Cementa.

[5]

Holm, G., Holmqvist, L., Johansson, S-E., Ljungkrantz, C., Retelius, A., Åhnberg, H.
(1995)

. Cement och kalk för djupstabilisering av jord, En kemisk - fysikalisk studie av stabi-

liseringseffekter, Statens geotekniska institut, SGI Rapport 48, Linköping.

[6]

Huttunen, Kujala, Vesa. (1996)

. Assessment of the quality of stabilized peat and clay, In-

ternational conference on ground improvement geosystems, 2, IS-Tokyo’96: Grouting and
mixing, Tokyo, May, 1996. Proceedinga, vol. 1, s 607–612.

[7]

Huttunen, Kujala, Lehto. (1996)

. Effect of humus on the binding reaction in stabilized

soils, International conference on ground improvement geosystems, 2, IS-Tokyo’96: Grout-
ing and mixing, Tokyo, May, 1996. Proceedinga, vol. 1, s 415-420.

[8]

Huttunen, Kujala. (1996)

. On the stabilization of organic soils, International conference on

ground improvement geosystems, 2, IS-Tokyo’96: Grouting and mixing, Tokyo, May, 1996.
Proceedinga, vol. 1, s 411–414.

[9]

Kujala, Ravaska. (1996)

. Settlement calculation of deep stabilized peat and clay, Interna-

tional conference on ground improvement geosystems, 2, IS-Tokyo’96: Grouting and mix-
ing, Tokyo, May, 1996. Proceedings, vol. 1, s 551–555.

[10]

Leppänen, M

. Masstabilisering av torv, Viatek, Esbo Finland, 1s.

[11] Rogbeck, Y., Sandin, P. (1995). Masstabilisering, Lägesrapport (arbetsmaterial), Statens ge-

otekniska institut, SGI projekt 1-9411-542, Linköping.

[12]

Rogbeck, Y., Berg, K-O., Säfström, L. (1996)

. Masstabilisering, Underlag till FoU-plan,

koncept, Statens geotekniska institut, Linköping.

References

Stabilization of Organic Soils

[13]

SGF (1995)

. Kalk- och kalkcementpelare, Vägledning för projektering, utförande och kon-

troll, Svenska geotekniska föreningen, SGF Rapport 4:95, Linköping.

[14]

TPPT (1995)

. Ground and surface stucture of road research project. Project 332. Deep stabi-

lization and masstabilization, Binder and material technology 332.20. Report on laboratory
investigations, Finska Vägverket (TPPT), Åbo.

[15]

Viatek (1994)

. Blockstabilisering av torv på Veittostesuo, Viatek Geoteknik, 2s.

[16] Personlig kommunikation med Elina Parkkinen Lohja Rudus Oy Ab.

[17] Byggvarudeklaration, yttre miljö – cement. Cementa AB.

Swedish Deep Stabilization Research Centre, Report 3

Stabilization of Organic Soils

APPENDIX 1

Swedish Deep Stabilization Research Centre, Report 3

APPENDIX 2.1

Stabilization of Organic Soils

APPENDIX 2.2

Swedish Deep Stabilization Research Centre, Report 3

APPENDIX 3

Stabilization of Organic Soils

APPENDIX 4

Swedish Deep Stabilization Research Centre, Report 3

APPENDIX 5

Stabilization of Organic Soils

APPENDIX 6

Swedish Deep Stabilization research Centre

c/o Swedish Geotechnical Institute, SE-581 93 Linkoping, Sweden

Phone: +46 13 20 18 61, Fax: +46 13 20 19 14.

Internet: www.swedgeo.se/sd

Document Outline