1999
Suction Caisson Installation in Shallow Water: Model Tests and Prediction
Installation de caissons à succion en eau peu profonde: essais et prédiction
Guo W.
School of Civil & Environmental Engineering, Nanyang Technological University, Singapore
Chu J.
Department Civil, Construction & Environmental Engineering, Iowa State University, USA
ABSTRACT: Suction caissons have been used as foundations to support mainly offshore structures such as offshore oil rigs in deep
water where a large suction pressure can be generated. Studies have been made recently to use this method for near shore foundations
in shallow water where the suction that can be applied is much smaller. In this paper, a study on the installation of suction caissons in
clay in shallow water using large scale model tests is presented. The model test setup and test results are discussed. The effects of soil
plug and side friction are evaluated. An analytical method proposed by Houlsby and Byrne is adopted to predict the penetration versus
time relationship. The analytical solutions agree well with the model test results.
RÉSUMÉ: Les caissons à succion ont été utilisés principalement pour les fondations de structures offshore en eau profonde
permettant de générer de fortes pressions de succion. Cet article présente une étude sur une installation de caisson dans de l’argile à
faible profondeur en utilisant un modèle à grande échelle. Les résistances d’arrachement et frottements latéraux sont évalués. La
méthode analytique proposée par Houlsby et Byrne est adoptée pour prédire la relation pénétration-temps et donne de bons résultats
KEYWORDS: Caisson; Clay; Model Test.
1 INTRODUCTION
A research project to use super-size cylindrical structures to
form underwater space and at the same time create land on top
is being carried out in Singapore. As the seabed soil is mainly
soft clay, suction caissons were considered on possible form of
foundations to support form part of the reclaimed land for
buildings or other types of structures to be built on top of it, the
foundation types for the offshore structures have to be
developed using innovative solutions. The most difficult design
condition is when the seabed soil is soft. It would be too costly
to treat the soft soil offshore. One innovative solution is to use
suction caissons.
Normally suction caissons are large, hollow, cylindrical steel
or concrete structures in form of upturned bucket shape, and are
penetrated into the seafloor bottom sediments by self-weight
and suction pressure. The principle of the suction caisson
technique is to apply suction inside a sealed cylindrical caisson
to create a downward net force to sink the caisson into the
seabed soil. After the suction is removed, the foundation is
constructed without treating the soft soil. The suction caisson
have been successfully employed in recent years in many
projects including mooring anchors (Andersen and Jostad,
1999; Andresen et al., 2011; Randolph et al., 2011; Wang et al.,
1975), beak water or sea walls (Chu et al., 2012), offshore
platforms (Zhang et al., 2007; Zhang and Ding, 2011) and
foundation for wind turbine in deep waters (Byrne et al., 2002;
Gavin et al., 2011; Houlsby et al., 2005c).
For caissons used in deep water, the hydrostatic water
pressure as provided by the water depth can contribute to
suction pressure to compress the caisson into seabed. However,
in relatively shallow water, there may not be sufficient suction
to allow the caisson to penetrate to the required depth. Another
factor affecting the penetration of a suction caisson is the soil
plug formed inside the caisson. When a caisson is penetrated
into clay, soil will go inside the open ended hollow caisson and
form soil plug. The soil plug resists the penetration of the
caisson. For this purpose and for the development of suitable
design methods, model tests and numerical studies were carried
out.
Analytical methods for analyzing the installation process of
suction caisson have also been proposed (Andersen et al., 2005;
Chen et al., 2009; Houlsby and Byrne, 2005a, 2005b; House
and Randolph, 2001; House et al., 1999; Tran and Randolph,
2008). In the method by Houlsby and Byrne (2005a, 2005b), a
constant penetration velocity was assumed. The driving forces
and soil resistance were also assumed to be balanced during the
whole installation process. This method was adopted to
calculate the amount of penetration of suction caisson subjected
to a constant driving force. The solution of this method was
compared with those from the model tests and good agreement
was achieved. Some of the key design parameters were also
evaluated based on the model test results.
2 MODEL TESTS
2.1 Soil Preparation
The soil used for the model tests was consolidated from kaolin
slurry. Factory made kaolin powder was used because of its
high coefficient of consolidation, low compressibility and
commercial availability. The kaolin used was supplied by
Kaolin Malaysia Sdn. Bhd. It has a specific gravity of 2.61, a
liquid limit of 61% and a plastic limit of 38%.
The kaolin powder was mixed with tap water into a slurry
form with water content of 81.3%. After mixing, the desired
slurry was transferred to the consolidation tank as shown in
Figure 1. Then the top cap and piston were mounted onto the
cylindrical tank. A compressed air pressure of 60
kPa
was
applied on top of the piston to consolidate the kaolin slurry for
about 10 days. The friction between the piston and the tank wall
was 17.35
kN
measured by a calibration test before the test.
Therefore, the effective consolidation pressure was 37.9
kPa
only.
The consolidated water was allowed to drain freely through a
drainage valve at the bottom of base plate. In order to
consolidate the kaolin slurry faster, a filter layer were designed
on the bottom of the tank including two layers of geotextile, fine
sand and gravel. The movement of piston was monitored by a
laser sensor (
Keyence
®
IL-600
). After the consolidation was
completed, the air pressure was reduced to zero and the top cap
was removed to allow soil samples to be taken for undrained
shear strength and water content tests. The water content of the
tested soil was 42.7%. The average undrained shear strength
