Sei sulla pagina 1di 41

6th Asian Regional Conference on Geosynthetics - Geosynthetics for

Infrastructure Development, 8-11 November 2016, New Delhi, India

GEOTEXTILE FILTERS AND


CONTAINMENT SYSTEMS FOR
RIVERBANK AND COASTAL EROSION
PROTECTION IN ASIA
T.W. Yee
Ten Cate Geosynthetics Asia, Malaysia

ABSTRACT
The mechanisms of erosion due to hydraulic action are discussed. A summary of conventional
riverbank and coastal erosion protection hydraulic structures is provided. This includes the
application of revetments, dykes and breakwaters. Rock or concrete systems are used as
armour units for the purpose of resisting the hydraulic forces of currents and waves. Before
the introduction of geotextiles granular materials are used in the bridging filter layers between
the armour units and the base soil that requires erosion protection. When geotextile filters were
introduced into the market in the 1960s they were used as replacement for one or more layers
of these granular bridging layers. The geotextile filter solution quickly gained acceptance due
to reasons of economics, ease of installation and better material quality control. Later on
geotextiles are used for the fabrication of containment systems to replace rock either as core
fill or armour units of such hydraulic structures. These geotextile containment systems are
generally filled with sand. However, some of the applications have involved the use of concrete
and other fill materials. The design of geotextile filters and geotextile containment systems
are presented. Geotextile containment systems were initially used generally for temporary
applications requiring service life of several years or way within permanent structures whereby
the exposure of these containment systems to the external environment spans only over the
construction period. Durability issues are discussed including the development of enhanced
durability containment systems that can withstand longer and more severe exposure conditions.
Several Asian case studies are presented.

1. INTRODUCTION
The Collins Dictionary of Geology (Lipidus 1990) provides a detailed definition of “erosion”. Erosion,
a noun, is the wearing away of any part of the Earth’s surface by natural agencies. These include mass
wasting and the action of waves, wind, streams and glaciers. Fundamental to the process of erosion is
that material must be picked up and carried away by such agents. Evidence for erosion is widespread;
the retreat of marine cliffs, deposition of fluvial material, and the cutting of great canyons, e.g. the
Grand Canyon. In order for erosion to occur three erosion mechanisms must act in sequence: detachment,
entrainment and transport. Detachment is caused by raindrop impact on a soil surface, running water
scouring action, flowing debris abrasion or wave breaking action. Entrainment is mainly due to fluid
drag. Once a particle is entrained, it tends to move as long as the velocity of the medium is high enough
to transport the particle horizontally. Depending on the impacting hydraulic energy, the landform present

KN - 210
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 211
Protection in Asia
and the characteristics of in-situ soil, erosion or deposition may occur at a specific location. Figure 1
shows the erosion and settling velocities for different soil particle sizes.

Fig. 1: Erosion and settling velocities for different soil particle sizes (after Pidwirny, 1999).
When we take a micro view of things, erosion then refers to the removal of soil particles from a specific
location and transportation to another location for deposition. When we take a macro view of things, in
particular riverbank and coastal erosion, such erosion events are evidenced by the loss of land to water
through a permanent retreat of the riverbank or shoreline. This paper is limited to the discussion on the
erosion of soil through hydraulic action; riverbank and coastal erosion; and the application of geotextile
filters and sand-filled geotextile containment systems for the control of erosion under such hydraulic
influence.
2. RIVERBANK AND COASTAL EROSION
2.1 Fluvial Processes and Riverbank Erosion
The hydrologic cycle involves the circulation and transformation of water throughout Earth’s atmosphere,
hydrosphere, lithosphere, and biosphere. Water evaporates when the Earth’s landmass and oceans is
warmed up as a consequence of exposure to radiation from the sun. Moisture condenses into precipitation
producing clouds in the atmosphere. Although most of the precipitation falls over the oceans, part of it
results in continental precipitation as a result of wind shifting of the clouds. Rivers form vast arterial
networks that drain the continents and in the process shape them. The term “fluvial” comes from the
Latin word “fluvius” which means “river”. Rivers produce fluvial erosion in which weathered material
and sediment on the riverbed is picked up for transport to new locations. Thus the river flow is a mixture
of water and its sediment load. Overland flow eroded material and collapsing riverbanks also contribute
to the sediment load in a river. Fluvial velocity and sediment load particle size generally determine the
rates of erosion and deposition. Sediment particles are deposited onto the riverbed at slower velocities
and are eroded at higher velocities.
Riverbank is defined as land at either edge of a river and riverbank erosion is the wearing away of the
bank of a river. The edge of a river or the contact line between the river and the land however is not a
static line. During high flow seasons the river widens and deepens as a consequence of higher discharge
and velocity but generally reverts to its previous low flow profile under stable fluvial regimes. Rivers
close to the sea are also subject to level changes under tidal influence. Riverbank erosion can manifest
itself in many ways e.g. progressive loss of land at the edge of a river, change in river bank profile usually
leading to an unstable slope, etc. River sections may be straight or curved, depending on a number of
factors. In curved sections deeper water and higher fluvial velocity tend to occur towards the concave
KN - 212 T.W. Yee

banks of the river. Therefore concave bends of river sections tend to be more prone to riverbank erosion
problems.
2.2 Littoral Processes and Coastal Erosion
Knowledge of the littoral process at a shoreline is necessary for a macro understanding of the depositional
or erosion framework of the shoreline. River sediment, material from collapsing sea cliffs and offshore
sand bars are some examples of littoral sources. The speed at which waves approach the shore depends
on sea floor and shoreline features and the depth of the water. As wave moves toward the shoreline,
different segments of the wave encounter the shoreline before others, which slows these segments down.
As a result, the wave tends to bend and conform to the general shape of the shoreline. Also, waves do not
typically reach the beach perfectly parallel to the shoreline. Rather, they arrive at a slight angle, called
the angle of wave approach. When a wave reaches a beach or shoreline, it releases a burst of energy
that generates a current, which runs parallel to the shoreline. This type of current is called longshore
current. Importantly, the longshore current not only moves water in the surf zone, it also moves sediment
parallel to the shoreline. Another sediment movement process is called beach drift. Sediment particles are
carried in the same direction that the waves are moving. Waves approaching at an angle to the shoreline
carry sediment upwards at the angle of wave approach.  When the water retreats, due to gravity it goes
downwards perpendicular to the shoreline, resulting in a beach drift in the same direction as that of
the longshore current. Sediment transported by beach drift as well as that moved by longshore current
together are called littoral transport.
Coastal erosion is defined as the wearing away of land and the removal of beach or dune sediment. The
shoreline is the contact line between the sea and the land. However, a shoreline is never static but shifts
with tides, storms and sea-level adjustments. A shoreline may retreat or advance during changing climatic
conditions within a year. Generally a stable shoreline experiences little changes over the years. Sediment
may be removed from a shoreline section during a specific season but replenished during other seasons.
Any disruption to a stable littoral process can however bring about permanent changes in a shoreline. For
example the imposition of new construction along the coast e.g. land reclamation into the sea, port and
harbour construction, breakwater construction, etc. can cause some sections of the shoreline to undergo
rapid erosion while other sections may be subject to shoreline accretion before the shoreline adapts and
stabilises to the new littoral transportation regime. Coastal erosion usually manifests itself as loss of land
to the sea due to the permanent retreat of the shoreline.
2.3 Erosion Hazard Assessment and Prioritisation for Implementation of Protection Works
Erosion hazard assessment studies may be carried out for specific rivers and coastlines or on a national
basis. These studies often help to identify areas and levels of erosion concerns. For example, the Hawaii
Coastal Zone Management Program (HCZMP 1993) identifies four levels of hazard severity ranking; i.e.
Low (1) for long term accreting shoreline with no history of erosion or dynamic cycles; Moderately Low
(2) for long term stable or minor erosion/accretion episodes; Moderately High (3) for long term erosion
rate of less than or equal to 0.33 m/yr or highly dynamic erosion/accretion cycles causing lateral shifts in
the waterline; and High (4) for chronic, long term erosion of greater than 0.33 m/yr or loss of beach and
seawall at water line during part of the tidal cycle.
Priority for implementation of protection works is usually based on risk of human and asset loss. For
example, in Malaysia 1400 km or 29% of the country’s coastline is facing erosion; of which 20% of
that is classified as Category 1 (or critical erosion whereby shore-based facilities and infrastructure are
under immediate danger of collapse or damage), 14% of that is classified as Category 2 (or significant
erosion whereby property and agriculture land are perceived as threatened within 5 to 10 years) and the
remaining 66% is classified as Category 3 (or acceptable erosion whereby it happens along an undeveloped
shoreline with minor or no consequent economic loss). Thus the priority for implementation of erosion
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 213
Protection in Asia
protection works will be the highest for Category 1 coastline in Malaysia. Any imposition of man-made
structures along natural riverbanks and coastlines not only requires erosion protection for the imposed
structure but their impact on the sediment and littoral transport need to be assessed for potential of
erosion downstream or adjacent coastline as a result of such impositions. Sediment and littoral transport
may be assessed by either physical or numerical modelling. For numerical modelling on sediment and
littoral transport examples of available programs include MIKE 21C (program that simulates river bed
and channel plan form development caused by changes in the hydraulic regime) by DHI; littoral transport
programs LITPACK by DHI and pyxisBeachplan by HR Wallingford; and DELFT3D 4 (program that
models sediment transport and bed dynamics) by Deltares. 
2.4 Riverbank and Coastal Erosion Protection Measures
To prevent erosion from occurring a variety of preventative measures are used to counter or reduce
the water forces acting on riverbanks and shorelines. Generally these measures fall into one of three
categories:
- Geometrical measures, where the bank or shore is altered in order to reduce the water forces below a
minimum threshold. Examples include provision and maintenance of a gentle sandy beach, etc.
- Stabilization measures, where the bank or shore is protected from erosion by stabilizing the susceptible
soil. Examples include the provision of revetments, etc.
- External measures, where the bank or shore is protected from erosion by the provision of protective
structures, placed at some distance in a manner that the littoral process is manipulated in a beneficial
manner. Examples include dykes and breakwaters, etc.
2.5 Revetments, Dykes and Breakwaters
A revetment is a protective cladding capable of withstanding hydraulic forces and used to protect the
sloping surface of an embankment or shoreline against erosion. A conventional revetment generally
consists of an armour layer that directly resists hydraulic forces but would usually include sub-layers of
bridging armour and filter layers. Dykes (reclamation dykes, containment dykes, sill dykes, river spur
dykes, coastal groynes, etc.) and breakwaters are generally broad based hydraulic structures with sides
sloping away from the crest. Beneath the armour layers of dykes and breakwaters there may be cores
and bedding layers of bridging and filter materials. The armour layers may be constructed of rocks,
concrete units, etc. Rock layers in revetments for erosion control applications may be in the form of
riprap, rock armour or stone pitching. Riprap is widely graded rock (D85/D15 ~ 2 – 2.5), which is placed
in bulk to give an armour layer about 2 to 3 stones thick. Pitched stone revetments are common in some
places where the increased stability afforded to relatively small rock by close packing has allowed the
use of local stone.
Other techniques that allow smaller stones to be used with improved stability include the use of steel
wire gabion and mattress products, and the use of cement grout, bitumen and other materials to bind the
stones. Rubble, which is usually rock or stone fragments, but may sometimes include broken concrete,
brick or asphalt, can be dumped to provide protection. Rock armour is more carefully selected rock of
a narrow size range (D85/D15 ~ 1.25 – 1.7.5), which is carefully placed in layers, usually two rock thick.
A wide variety of precast concrete products are available for use to replace of rock as erosion control
cover units (Przedwojski et al. 1995, McConnell 1998, Pilarczyk 2000). They either come in the form of
block or slab units. For coastal and marine structures exposed to high intensity wave attack rock armour
of sufficient weight and size is often used. Alternatively, precast concrete armour units e.g cube, tripod,
tetrapod, tribar, dolos, stabit, etc. may be used to replace rock armour which may be unavailable or non-
economical at the project site.
KN - 214 T.W. Yee

3. GEOTEXTILE FILTERS FOR EROSION PROTECTION STRUCTURES


The use of geotextile filters in marine and hydraulic engineering has a long and successful history
(Heibaum et al 2006). Geotextiles began to be used as filters in coastal protection structures in Asia in
the mid 1970’s following on from earlier experiences in Europe and North America; with Japan and
Singapore being early adopters of these materials (Lawson 2012). At that time, little was known about
the performance of geotextile filters in a hydraulic environment and thus, it was the more “adventurous”
(today we would probably describe them as “innovative”) authorities that adopted this “new” technology.
While there were some disappointments (mostly due to incorrect installation) the vast majority of
applications were considered very successful and by the mid 1980’s the use of geotextile filters had
spread through much of Asia.
3.1 Classification of Geotextiles
Many different types of geotextiles have been developed for use in filtration applications. Geotextiles
are generally made from synthetic fibres rather than natural fibres. The synthetic materials, or
polymers, are made in chemical processing plants from the polymerization of thermoplastics.
Geotextiles are commonly made from the polymers of polypropylene (PP), polyester (PET) or
polyethylene (PE). Although there is a wide variety of geotextiles, they generally fall into one
of three categories, nonwoven, woven or composite. There are two different fibre types used to
manufacture nonwoven geotextiles: continuous filaments; and staple fibres. Both fibre types are
made from continuously extruded circular cross-section filaments (Figure 2a). The only difference
between the fibre types is their length. Continuous filaments are fibres of extreme length and staple
fibres are very short fibres (in the range of 25 - 100 mm long). Nonwoven geotextiles (Figures 3a,
3b) consist of either continuous filaments or staple fibres randomly distributed and bonded either
through a needlepunching process or a heat bonding process.
There are four main yarn types which are used to manufacture woven geotextiles: multifilament
yarn, slit-tape yarn, fibrillated tape yarn and monofilament yarn. Their diameters are the same as
the diameters of the holes in the spinneret, unless they are altered through a drawing or stretching
process. Multifilament fibres (Figure 2b) are formed from yarns which consist of many fibres. The
diameters of the fibres depend on the number of monofilament fibres used, and how they are combined
to form a yarn. Slit-film fibres (Figure 2c) are made from continuously extruded flat tapes that are cut
into narrow strips by knives or air jets. Fibrillated fibres (Figure 2d) are made from extruded films,
where short, discontinuous cuts have been made down their length. Monofilament yarns (Figure
2e) are fibres that are manufactured from circular and quite often oblong cross-section filaments.
Woven geotextiles (Figures 3c, 3d, 3e) consist of fibres or yarns of a polymer that are oriented in two
perpendicular directions, one over the other.

Fig. 2: Types of fibers and yarns typically used for manufacture of geotextiles (a) filament yarn (b)
multifilament yarn (c) slit tape yarn (d) fibrillated yarn (e) monofilament yarn (adapted from Bhatia and
Smith 1996)
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 215
Protection in Asia

Fig. 3: Types of geotextiles (a) needle punched nonwoven (a) heat bonded nonwoven
(c) woven multifilament (d) woven slit tape (e) woven mono filament
(adapted from Bhatia and Smith 1996; Aydilek and Edil 2002)
3.2 Filtration Mechanisms
The mechanisms by which particles are retained by fabric media may be identified as cake filtration,
screening and depth filtration (Hardman 2000). Cake filtration involves the accumulation of particles that
bridge together in a porous structure on the surface of the fabric. It follows from this that, once formed,
the cake effectively becomes the filter medium with the fabric thereafter acting simply as a support.
Screening (also known as straining) is a simple mechanism in which particles are confronted with an
aperture which is smaller than the particles themselves. Depth filtration is the mechanism where particles
are captured through attachment to the fibres within the body of the filter medium, e.g. because of Van der
Waal or electrostatic forces, even though they may be smaller than the apertures that are formed. Depth
filtration is particularly relevant to thick needlepunched nonwoven geotextiles. Figure 4 illustrates the
three different filtration mechanisms.

(a) Cake filtration


Flow of liquid

Flow of liquid

(b) Screening

(c) Depth filtration


Fig. 4: Filtration mechanisms (a) cake filtration (a) screening (c) depth filtration (Hardman 2000)
3.3 Geotextile Filters
3.3.1 Revetment Filters
Geotextile filters are “sheet” materials which makes them intrinsically hydraulically stable (they cannot
be washed out through the granular layer above, unlike granular filters). In Asia today geotextiles have
KN - 216 T.W. Yee

virtually replaced the use of graded granular filters in revetment constructions for cost, engineering
effectiveness, installation expediency and other reasons. Figures 5a and 5b shows a traditional revetment
structure constructed with granular filters and a commonly used revetment structure of today constructed
with geotextile filter respectively. Well established procedures exist for the design of revetments
incorporating granular materials. First, the outer, primary armour layer must remain stable under the
design water forces acting on the revetment. Second, the under-layers in the revetment must be designed
to be hydraulically stable (i.e. an under-layer cannot be washed out through the layer above). Third, the
filter layer must remain intact with the soil it is protecting and also be hydraulically stable (as for the
rest of the under-layers). CIRIA (2007) provides comprehensive design guidance for the sizing of rock
armour and granular under-layers for revetments.
Geotextile filters are placed from the top of the graded slope of a revetment structure. The roll of geotextile
is unwound downwards from the top of the slope to the bottom toe and secured with either sandbags,
stones or pegs placed by divers. The installation of the geotextile filter may be carried out using land-based
machinery sitting on top of the slope or by water waterborne placement with the help of a pontoon.

(a) (b)
Fig. 5: Rock armour revetment (a) constructed with granular filters (b) constructed with geotextile filter

3.3.2 Basal Filters in Dykes and Breakwaters


Where rubble-mound dykes and breakwaters are located on rock or over consolidated clay foundations
erosion across the base and at the toes of the dyke or breakwater generally does not occur. However, if
the dyke or breakwater is located on a sand foundation then erosion and scour at the toe of the dyke or
breakwater and across its base can occur which may lead to undermining and instability. In these situations
a geotextile filter, located across the base of the rubble-mound breakwater, is included to prevent erosion
and scour of the sand foundation. For reclamation and containment dykes, soil is placed on one side of
the dyke and a geotextile filter is also placed in between the dyke and the infill soil. The engineering basis
for design of dykes and breakwaters is the same as that for design of revetments. First, the outer, primary
armour layer must remain stable under the design water forces acting on the dyke or breakwater. Second,
the under-layers and core of the dyke or breakwater must be designed to be hydraulically stable (i.e. an
under-layer cannot be washed out through the layer above). Unlike revetments, dykes and breakwaters
make contact with soil at its base. Thus, the filter layer is laid on the bed rather than on slope. Similarly,
the filter layer must remain intact with the base soil it is protecting and also be hydraulically stable (as
for the rest of the under-layers and core). Figures 6a and 6b shows a traditional rubble mound dyke or
breakwater structure constructed with basal granular filter and a commonly used rubble mound dyke or
breakwater structure of today constructed with basal geotextile filter respectively.
Depending on the specific location of the dyke or breakwater the basal geotextile filter may need to
be installed at appreciable water depth and in diverse water current and wave conditions. To facilitate
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 217
Protection in Asia
this, the basal geotextile filter is normally fabricated on-land into what is known as a “fascine mattress”
prior to floating to site and sinking into place. Historically, this technique was originally developed in
Europe using “fascines” (brushwood) to support the geotextile filter during flotation and installation. This
technique has been used extensively in Asia since the early 1980’s, however, bamboo has been substituted
for the fascines as it is more-plentiful, light and very strong. The fascine mattress technique involves the
on-land fabrication of large geotextile sheets attached to a lattice bamboo network. The “mattress” is then
pulled into the water and floated to its installation location and then sunk into place on the seabed by the
application of rock on top of the mattress. Figures 7a and 7b show the application and typical installation
of revetment filters and basal filters for marine dykes and breakwaters respectively.

(a) (b)
Fig. 6: Dyke or breakwater (a) constructed with basal granular filter
(b) constructed with basal geotextile filter

(a) (b)
Fig. 7: Application and typical installation (a) of revetments filters
(b) of basal filters for marine dykes and breakwaters

4. GEOTEXTILE CONTAINMENT SYSTEMS FOR EROSION PROTECTION


STRUCTURES
Geotextile containment systems first began to be used for hydraulic and marine structures in the 1960s
(Lawson 2008). Geotextile containment systems for riverbank and coastal protection were first introduced
into Asia in the early 1990s. In fact, Asia has since been at the forefront of many innovative uses of
geotextile containment systems in riverbank and coastal protection applications. In Asia mainly three
types of geotextile containment systems are used for erosion protection works. These are geotextile bags,
geotextile mattresses and geotextile tubes. In Asia geotextile bags and mattresses are mostly used for the
construction of revetments while geotextile tubes which tend to be significantly larger and heavier units
are mainly used for construction of dyke structures. By and large the geotextile containment systems used
in Asia mostly involve sand as fill material although grout filled geotextile mattresses are also used.
KN - 218 T.W. Yee

4.1 Geotextile Bags


Geotextile bags are manufactured in a range of sizes and shapes, ranging in volume from 0.3 m3 to around
8 m3, and may be pillow-shaped, box-shaped or mattress-shaped. When used for revetments, an important
feature of geotextile bags is that they are installed in a pattern-placed arrangement that greatly improves
their overall stability and performance. Geotextile bags are factory fabricated and are delivered to site
where they are filled. When designing geotextile bags due consideration must be given to the mechanical
stresses acting on the bags during sand-filling, placement and throughout the life of the bag structure.
Normally, the maximum mechanical stresses occur during sand-filling and bag placement. Sand-fill is
used within the geotextile bags; however, in specific situations other fill types may be used. Sand-fill
provides good stability when in a confined state and does not change in volume over time. While it is
not imperative to fill the geotextile bags to maximum fill volume, for maximum stability this should be
done.
Filling to maximum volume reduces the ability of the sand-fill to move inside the geotextile bags when
subject to hydraulic forces and thus minimizes localized bag arching and hence premature instability. To
maximize filling volume it is common practice to introduce water during the sand filling procedure to
maximize the contained density of the sand-fill in the geotextile bags. Figure 8a shows the mechanical
filling of a geotextile bag with sand prior to placement in position. Figure 8b shows a stream channel
protected with geotextile bag revetment structure.

(a) (b)
Fig. 8: Geotextile bag (a) mechanically filled with sand (b) used as a revetment.
4.2 Geotextile Mattresses
Geotextile mattresses consist of two connected layers of geotextile where the space between is filled
generally with sand. The resulting mattress shape is composed of sand-filled tubular compartments that
maintains their mass-gravity and prevents movement of the sand-fill within the mattress. Geotextile
mattresses are manufactured using a woven geotextile backing with a composite geotextile facing. The
backing provides the primary strength requirements dependent on the method of installation (e.g. if the
mattress is laid on the slope and then filled, or pre-filled and then placed on the slope). Commonly, the
facing is a composite geotextile that contributes strength in order to contain the sand-fill, traps sediments to
promote vegetation growth and has good durability properties (i.e. good resistance to UV light exposure).
The prefabricated geotextile mattress is hydraulically filled with sand. The resulting mattress weight and
thickness, and consequently, its stability is related to the spacing between the sewn seams of the tubular
shapes in the mattress. However, to prevent future movement of the sand-fill within the mattress the
spacing between these sewn seams should be limited. Consequently, average filled geotextile mattress
thicknesses range from 0.1 to 0.2 m. Performance-wise, geotextile mattresses lie between that of Erosion
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 219
Protection in Asia
Control Mats and the more substantial geotextile bags with regard to the level of erosion protection
stability provided. Figure 9a shows the hydraulic filling of a laid in position geotextile mattress with sand.
Figure 9b shows a river channel protected with geotextile mattress revetment structure.

(a) (b)
Fig. 9: Geotextile mattress (a) hydraulically filled with sand (b) used as a revetment

4.3 Geotextile Tubes


Geotextile tube is defined as “a large tube [greater than 7.5 feet (2.3 m) in circumference] fabricated
from high strength, woven geotextile, in lengths greater than 20 linear feet (6.1 m)”, according to GRI
Test Method GT11: Standard Practice for “Installation of Geotextile Tubes used as Coastal and Riverine
Structures”. Geotextile tubes used in coastal and riverine applications are typically filled hydraulically
with sand. Also associated are “filling ports” which are geotextile sleeves sewn into the top of the
geotextile tube into which the discharge pipe is inserted. They are typically laid at the final intended
location prior to filling. Initially, the filling ports at the extreme ends of the geotextile tube are utilised
while those in-between are temporarily closed. One end is for the pumping in of sand slurry while the
other end is for water pressure relief and discharge. In this way, the sand slurry will flow from one end to
the other of the geotextile tube and gradually depositing sand along the way. Figure 10 shows a schematic
of the hydraulic filling of a geotextile tube with sand. It may be necessary to move the filling point in
order to achieve more even filling of the geotextile tube. After completion of filling the port sleeves are
closed and attached to the geotextile tube in a manner sufficient to prevent sand loss and movement of the
sleeve by wave action. Figure 11 shows the application of geotextile tube as hydraulic structures shown in
3-D with cut-away sections. Figure 12a shows the use of geotextile tube as coastal dyke in a permanently
exposed form. Figure 12b shows the use of geotextile tube as the core of a rock dyke.

Fig. 10: Schematic of the hydraulic filling of a geotextile tube with sand
KN - 220 T.W. Yee

(a)

(b) (c)
Fig. 11: Application of geotextile tube (a) in dyke or breakwater (b) to retain perched beach
(c) as reclamation berms

(a) (b)
Fig. 12: Geotextile tube (a) as exposed coastal dyke (b) as covered core of a rock dyke.

5. DESIGNING WITH GEOTEXTILE FILTERS


There are three basic filter criteria which are used for the selection of a geotextile as a filter: (i) a retention
requirement, to prevent the migration of soil particles through the geotextile; (ii) a permeability requirement,
to ensure the free flow of liquid through the geotextile; and (iii) a non-clogging requirement, to ensure the
geotextile will adequately meet the permeability and retention criteria throughout the life of the structure
(Bhatia and Smith 1996). Quite a few filtration criteria have been developed empirically (Carroll 1983,
Giroud 1996, Pilarczyk 2000). The geotextile filter must also survive the installation process, without
suffering mechanical damage. Therefore, the geotextile installation survivability requirement form an
important part of the geotextile filter design process.
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 221
Protection in Asia
5.1 Particle Retention Requirement
The recommendation by Pilarczyk (2000) developed for stable soils may be used to check the particle
retention criteria for design of geotextile filters in revetments, dykes and breakwaters. These criteria have
been developed for application specific to geotextile filters under current and wave attack.
Table 1: Particle retention criteria (adapted from Pilarczyk, 2000)
D40< 60 µm D40 > 60 µm
Hydraulic load Requirement 1 Requirement 2 Requirement 1 Requirement 2
Stationary load (current)

Dynamic load (wave) µm

O90 = pore size of the geotextile filter;


D10 = sieve size through which 10% fraction of the soil material passes;
D40 = sieve size through which 40% fraction of the soil material passes;
D60 = sieve size through which 60% fraction of the soil material passes;
D90 = sieve size through which 90% fraction of the soil material passes;
Cu = uniformity coefficient (=D60 / D10)
5.2 Permeability Requirement
The geotextile permeability criteria require the geotextile permeability to be higher than a factored base
soil permeability and take the form of a generalised equation as follows:
kg > fp ks [1]
where, kg = permeability of the geotextile filter; fp = partial factor applied for permeability; and ks =
permeability of the base soil. Pilarczyk (2000) recommends a value of fp = 2 if the base soil is clean sand
but suggested a higher value may be necessary otherwise. McConnell (1998) recommends a general
value of fp = 5 for design. Other authors have recommended different values for fp that range between 1
to 10.
5.3 Non-Clogging Requirement
Distinction should be made between clogging of filter cake and clogging of geotextile filter. Clogging
of filter cake is only an issue when the filtration mechanism developed is cake filtration. Clogging of
geotextile filter or sometimes referred to as internal clogging can happen for all filtration mechanisms,
depending on hydraulic conditions, soil environment and nature of geotextile filter. Clogging of geotextile
can be simply defined as reduction in pores and/or channels within the geotextile to an extent that the
reduced permeability of the geotextile begins to compromise its performance as a filter medium. Clogging
of geotextile can occur in one or more of three methods i.e. physical, chemical and biological clogging.
Physical clogging is a result of soil particles closing up pores on the surface of the geotextile or getting
trapped within the internal structure of the geotextile. The risk of physical clogging of thin geotextiles
is very low and is a risk only with thick nonwoven geotextiles where the depth filtration mechanism
dominates. Chemical clogging of geotextiles occur when dissolved solids precipitate out when exposed to
oxygen in the air or otherwise and deposit onto and into the geotextile. Biological clogging of geotextile
occurs when biological growth happens on the surface or inside the geotextile. Biological clogging is
usually only an issue in aggressive biological environment e.g. leachate filtration in landfills but is seldom
a problem in the hydraulic and marine environment.
KN - 222 T.W. Yee

Existing clogging criteria are generally based on: soil/geotextile filtration tests; geotextile porosity
measurements (Christopher and Holtz 1985); or on relationships between a characteristic pore size of
the geotextile, such as O95, O90, O50 or O15 (Christopher and Holtz 1985; Fischer et al. 1990) and the
grain-size distribution of the surrounding soil. The clogging potential of a geotextile may be evaluated
by performance tests, such as the Long-Term Flow Test (Koerner and Ko 1982), the Gradient Ratio
Test (ASTM D 5101), or the Hydraulic Conductivity Test (ASTM D 5567). These performance tests
are generally not carried out because they are time consuming, expensive, and only provide information
for a specific soil/geotextile system. The porosity of a geotextile is important; however, almost every
geotextile passes the given criterion that a geotextile should have a porosity value greater than 30-40%
and thus may not be truly reflective of its clogging potential. Non-clogging requirement is thus evaluated
by considering the permeability and the opening size of the geotextile. Effectively, when a significantly
high value of fp (say 3 or higher) is used, it becomes an insurance against clogging.
Traditionally, nonwoven geotextiles are manufactured using continuous or staple fibres of a single diameter
for a specific product range of nonwovens. When higher mechanical strength is required, more passes
of fibre laydown is effected before the bonding process begins. When nonwoven is used as geotextile
filter, to survive the placement of rock on top of the filter layer, the nonwoven required usually ends
up quite thick. This is especially so when less bridging layers are used in design requiring larger stones
to be dropped directly onto the geotextile filter. This led to the development of a two-layer composite
nonwoven geotextile filter product (Giroud et al 1998, Delmas et al 2000). Filtration is functionally
provided by a layer of needle-punched nonwoven of specific thickness constructed with smaller diameter
fibres (facing the base soil) and mechanical survivability is functionally provided by a variable thickness
layer of needle-punched nonwoven constructed with larger diameter fibres (facing the rock drop). The
filtration functional layer provides the optimum properties for particle retention while the larger diameter
fibre layer can be varied to provide the necessary mechanical properties to survive the rock drop process.
By using larger diameter fibres to add thickness of the survivability functionally layer the higher porosity
within this layer helps ensure low clogging risk irrespective of how much fibres is added to achieve the
necessary mechanical properties to survive the rock drop process. The two layers are identified usually
through the use of different fibre colour for the smaller and larger diameter fibres.
Giroud (1996) developed a theoretical model for nonwoven geotextile filters and determine the minimum
useful thickness of the filter. A nonwoven is made of fibres randomly laid in space and a soil particle which
passes through the nonwoven geotextile filter must pass through several passages, called constrictions,
delimitated by three or more fibres crossing one another. Giroud et al (1998) established relationships
between the constriction size distribution curve and the opening size distribution curve of the nonwoven
geotextile filter. Giroud et al (1998) also recommended that for optimal filtration performance and to
satisfy the non-clogging criterion, the filtration functional layer of the two layered nonwoven geotextile
filter should have pore size from 40 microns to 80 microns and constrictions numbered from 25 to 40.
5.4 Installation Survivability Requirement
Figure 7a shows the placement of rock directly over the installed geotextile filter in a coastal revetment.
The geotextile filter in the revetment structure must be robust enough to resist any mechanical stresses
applied to it during installation, otherwise called the installation survivability requirement of the geotextile
filter. Normally, the critical mechanical stresses occur during the placement of the rock layer(s) above the
geotextile filter. Very early on in the use of geotextiles in revetments in Asia it was recognised that it was
critical for the geotextile filter to resist puncturing and tearing during the placement of the rock layer(s)
on top. Lawson (1982) developed an empirical relationship between the required geotextile dynamic
puncture resistance and the rock being placed on top, based on the evaluation of a number of sites in
Asia. Following further evaluation of revetment sites in Asia, Lawson (1992) was able to present a more
universal empirical relationship for the geotextile puncture resistance requirement, being:
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 223
Protection in Asia

MA > 1200 D85 ඥ݄௢ [2]


The use of the geotextile mass per unit area parameter as the determinator of puncture resistance enabled
the empirical relationship to be used for a much wider range of rock sizes than the original relationship.
This relationship was derived on the basis of rock drop on geotextile in air but may be generalised for rock
drop that involves falling through water as well if the equation is rewritten in terms of impacting velocity
of rock on geotextile. Using energy conservation principles Equation 2 may be rewritten as follows (as
the rock drop design height through air is related to the impacting velocity of rock on geotextile):
ܷௗ

MA > 1200 D85 ඥʹ‰ [3]


where, MA = mass per unit area of the geotextile filter (in g/m2) measured according to ISO 9864:2005;
D85 = 85% rock size on geotextile (in m); ho = height of rock drop on geotextile in air (in m); Ud = the
design velocity of the stone impacting the geotextile (in m/s); and = gravitational acceleration (in
m/s2). To generalise the situation to cover rock drop that involves falling through water as well (shown
in Figure 13), the equations developed by Berendsen (1996) may be used to determine the generalised
impacting velocity of rock on geotextile and are given in a modified form below:
ඥܷ௘ଶ ൅ ሺܷ௢ଶ െ ܷ௘ଶ ሻ݁ ିଶ௖భ ௛ೢ
Uw = [4]
ହሺఘೞ ିఘೢ ሻ୥஽ఴఱ

Ue = ଷఘೢ ஼೏
[5]
͵ߩ௪ ‫ܥ‬ௗ

c1 = ͷߩ௦ ‫଼ܦ‬ହ [6]
where, Uw = the velocity of the rock at depth of hw in water (in m/s); Ue = the equilibrium velocity of the
rock in the water (in m/s); Uo = the velocity of the rock when hitting the water (in m/s); hw = the water
depth of the bed (in m); ρs = the density of rock (in kg/m3); ρw = the density of water (in kg/m3); =
gravitational acceleration (in m/s2); D85 = 85% rock size on geotextile (in m) and Cd = the drag coefficient
of the stone (-).
For basal filters the geotextile is placed on the bed and therefore the rock fall through water before
impacting with the geotextile is constant at hw; and thus Ud = Uw. For revetment filters the geotextile is
placed on ground, over the revetment slope and on the bed at the toe of the revetment; and therefore the
rock fall through water before impacting with the geotextile can vary from 0 to hw. For revetment filters
two limiting conditions may exist. When Uo > Ue, the velocity of the falling rock in water will decrease
until the value of Ue is reached. The largest possible rock impact velocity then is Uo; and thus Ud = Uo.
When Uo < Ue, the velocity of the falling rock in water will increase until the value of Ue is reached. The
largest possible rock impact velocity then is when the rock impacts geotextile on the bed at depth hw; and
thus Ud = Uw. Once the appropriate value of Ud is determined, this value may be substituted into Equation
3 for the geotextile selection to satisfy rock drop survivability.

Fig. 13: Generalised situation involving rock drop from a datum of constant rock drop height through air.
KN - 224 T.W. Yee

Watn and Chew (2002) discussed the issue of rock drop survivability of geotextile filters in riverbank and
coastal protection works covering theoretical approaches and field testing in some detail. The theoretical
approaches included the energy concept for damage and the energy balance concept for assessing a
geotextile filter’s resistance to rock drop. Typically, many countries adopt local experience in developing
the rock drop survivability criterion. However, these tend to be limited to general classifications based
on specific index tests. For example, the BAW RPG 3.10 - Impact Test according to Federal Waterways
Engineering and Research Institute– Guidelines for Testing Geotextiles for Navigable Waterways is used
to quantify drop energy and survivability in Germany as outlined in Federal Waterways Engineering and
Research Institute – Code of Practice on Use of Geotextile Filters on Waterways (BAW 1993, BAW 1994,
Heibaum 2014) but the testing capability outside of Germany rarely exists. On the other hand, AASHTO
M288 in the United States provides selection guidance based on universal index tests but the application
scope may be limiting. As a general guideline AASHTO M288 does not permit rock drop height of
greater than 1 m. The universal empirical relationship based on mass per unit area of the geotextile
filter by Lawson (1992) given as Equation 2 and rewritten as Equation 3 provides a convenient tool for
prequalification and cost budgeting of geotextile filters for riverbank and coastal protection works. The
empirical relationship by Lawson (1992) can be combined with experience-based impact strength and
energy balance concepts to derive material specification that is based on a set of universal index test
properties.
More recently, Chew et al. (1999) and Wong et al. (2002) carried out quite extreme rock installation damage
tests (large rocks dropped from up to 5 m height) to arrive at suitable geotextile mechanical properties
that are now used as the basis for coastal protection structures in Singapore. Two interesting observations
were made during this research program. First, having a soft foundation beneath the geotextile filter was
beneficial in its resistance to installation damage by rocks (compared to a firm foundation). Second, the
dropping of large rocks onto a stone bedding layer above the geotextile filter resulted in greater damage
to the geotextile than the installation of the bedding layer itself. Good practice when placing rock layers
above the geotextile filter would be to limit the height of placement to a maximum of 1.5 m. This has the
dual benefit of limiting the puncture stresses that cause damage to the geotextile filter and ensuring better
consistency and accuracy of rock placement in the revetment structure.
The deployment and installation of basal filters normally impart quite high tensile loads to the geotextile
filter. The forces may result from current forces acting on the fascine mattress or incidental pulling on
the free end, when the fascine mattress has been anchored down at one end with some initial ballast
weight. Consequently, woven geotextiles with tensile strengths ranging from 100 kN/m to 200 kN/m are
commonly used for this application. However, it should be noted that higher strength woven geotextiles
have also been used. Thus in addition to checking for survivability against rock drop, basal filters need to
be checked for tensile strength needed to survive deployment and installation. The critical situation arise
when one end of the fascine mattress basal filter has been held down with the initial rock ballast while
the other end is still held up at water surface and the fascine mattress basal filter is subject to drag forces
induced by current impacting on it.
6. DESIGNING WITH GEOTEXTILE CONTAINMENT SYSTEMS
Bezuijen & Vastenburg (2013) provide comprehensive design guidance for geotextile containment systems
in erosion protection applications. The design life of the hydraulic structure should be defined. This is
most likely determined on an economic basis and this definition impacts on the product specification for
durability requirements. The required level of service of the revetment, dyke or breakwater should also be
determined. This may be in terms of an acceptable overtopping limit or risk of damage. Last but not least
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 225
Protection in Asia
the design event of a certain return period should be selected. The definition of geometry is necessary for
assessment of both hydraulic and geotechnical stabilities. The definition of geometry would include slope
angle, crest elevation, toe levels, etc. Typically, the information required as design input include water
level, wave conditions, wave direction and current flow conditions. Design water levels should include
the influence of tides and surges if applicable. The design water level is also referred to as still water level
(SWL). SWL is the level that the water surface would assume in the absence of wind waves. The design
of revetment with sand filled geotextile bags and mattresses is discussed. The design of dyke structure
using sand filled geotextile tubes is also discussed.
From a technical standpoint the geotextile containment systems need to fulfil the following:
• Internal stability
o The geotextile used to fabricate the geotextile containment unit, including seams and closure,
need to withstand the stresses encountered during the installation process (commonly referred to
as the tensile strength requirement).
o The geotextile should prevent loss of fines during installation and under in-service wave and flow
attacks (commonly referred to as the sand tightness requirement).
• External stability
o The sand filled geotextile containment structure should be stable against wave and current attacks
(commonly referred to as the hydraulic stability requirements).
o The sand filled geotextile containment structure should be stable against sliding, overturning, bearing
and global slip failures (commonly referred to as the geotechnical stability requirements).
• Durability
o The geotextile used to fabricate the geotextile containment unit need to last over the design life of
the structure (commonly referred to as the durability requirement). This includes durability of the
polymer material due to chemical degradation in a buried state as well as durability against UV,
abrasion, mechanical damage from impacting debris, etc. during the state of exposure.
6.1 Tensile Strength Requirement
The tensile strength requirement refers to the ability of the geotextile containment system to resist rupture
or failure during the installation process. This forms part of the mechanical strength requirement required
over the design service life which includes mechanical properties to survive the exposure duration to the
environment e.g. abrasion, incidental mechanical impacts, etc. and will be discussed later under durability
issues.
6.1.1 For Geotextile Bag Applications
The tensile force induced during the filling of a geotextile bag is rarely critical. The tensile strength of
the fabric, seams and lifting straps required to survive the handling and placement process depend on the
weight of the filled geotextile bag. If a geotextile bag is dumped under water, it must be robust enough
to survive impact with the bottom. Bezuijen and Vastenburg (2013) proposed a tensile strength check as
follows:
‫ܦ‬௘ ܸ ‫ܬ‬௚ U௙ െ U௪ [7]
ܶ௚ ൒ ඨʹ U௙ ‰
ܾܵ ‫ܥ‬ௗ U௪

KN - 226 T.W. Yee

where, Tg = required tensile strength of the geotextile bag (in kN/m); De = effective thickness of the
sand-filled geotextile bag (in m); V = volume of the sand-filled geotextile bag (in m3); Jg = tensile
stiffness of fabric of geotextile bag (in kN/m); ρf = bulk density of sand-fill in geotextile bag (in kg/
m3); ρw = density of water (in kg/m3); g = gravitational acceleration (in m/s2); b = width of the sand-
filled geotextile bag (in m); S = circumference of the sand-filled geotextile bag (in m); and Cd TP = the
drag coefficient (-).
6.1.2 For Geotextile Mattress Applications
Tensile forces may be induced in the geotextile mattress through the following manner i.e. during filling
with sand, in the trenching process and in the hoisting of the pre-filled geotextile mattress (if pre-filling
is done before placement in position). The tensile forces are rarely critical except in the case of hoisting
of a pre-filled geotextile mattress. Bezuijen and Vastenburg (2013) proposed a tensile strength check as
follows:
݈ܾ௙ ‫ܦ‬௘
ܶ௚ ൒ U ‰ [8]
ܾ௨ ௙

where, Tg = required tensile strength of the geotextile mattress (in kN/m); l = length of the sand-filled
geotextile mattress (in m); bf = width of the sand-filled geotextile mattress (in m); De = effective
thickness of the sand-filled geotextile mattress (in m); ρf = bulk density of sand-fill in geotextile
bag (in kg/m3); g = gravitational acceleration (in m/s2); and bu = width of the flat unfilled geotextile
mattress (in m).
6.1.3 For Geotextile Tube Applications
The analytical method described by Timoshenko (1959) is based on the equilibrium of an encapsulating
flexible shell filled with pressurized slurry, resulting in the tensile force being constant over the entire
circumference of the geotextile tube. This method is used for various design softwares (GeoCoPS,
SOFFTWIN, Deltares, Geotube® Simulator, etc.) that will determine the tensile force and geometry of
the geotextile tube (Leschinsky & Leshchinsky, 1995; Leshchinsky et al, 1996; Palmerton, 1998, 2002;
Bezuijen & Vastenburg, 2013; Plaut & Suherman, 1998). They are typically based on the assumptions
that the geotextile tubes are made from an inextensible fabric and installed over an unyielding flat
ground surface. Some analytical formulations can account for yielding ground surface (Kazimierowicz,
1994; Plaut & Suherman, 1998; Guo et al, 2011) while some analytical formulations take into account
extensibility of the geotextile tube (Cantré & Saathoff, 2011; Yee, 2012).The fabric tensile strength
requirement is typically specified based on the factored tensile force. Typically, a high global factor of
4 to 5 to account for construction damage, environmental damage, seam efficiency, etc. is not unusual.
Besides, fabric tensions in the field have found to be higher than analytical predictions using conventional
fully fluid contained model (Yee 2016, Yee et al 2010). Creep is rarely a major concern even when
polypropylene is used. Upon cessation of the filling process, the fabric tension will drop as a result
consolidation of tube height and a changeover of contained fill material from a fully fluid state to a solid
state. That fact and the high factor of safety typically adopted in design will result in residual sustained
stresses at levels where creep effects are negligible. The charts in Figure 14 may be used for preliminary
designs in the absence of design software.
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 227
Protection in Asia

(a) (b)

(c) (d)
Fig. 14: Geotextile tube design (a) diagram showing terminology (b) chart relating tension factor, TF to HT/DT
(c) chart relating BT/DT and bT/DT to HT/DT (c) chart relating AT/DT and bT/DT to HT/DT

The working circumferential fabric tension of the filled geotextile tube, Tc is related to the tension factor,
TF determined from Figure 14b according to the equation below:

ܶ௖ ൌ ͲǤͷߛ௙ ‫ ்ܪ‬ଶ ሺܶி ൅ ͲǤͷሻ [9]


where, TcTP = working circumferential fabric tension of the filled geotextile tube (in kN/m); γf = unit
weight of sand-fill in geotextile tube (in kN/m3); HTTP = geotextile tube filled height (in m); and TFTP =
tension factor derived from Figure 14b (-). As an illustration, if a geotextile tube with DT = 4 m is filled
KN - 228 T.W. Yee

to HTTP = 2 m using sand-fill with γf = 19 kN/m3, then HTTP/DT = 0.5. Consequently, FTTP = 0.085 (see
Figure 14b) which when substituted into Equation 7 will give a value of TcTP = 22.2 kN/m. From Figure
14c, BT /DT = 1.32 and bT /DT = 1.05; which results in the geotextile tube width, BT = 5.3 m and the contact
width, bT = 4.2 m. From Figure 14d, AT /DT = 0.56; which results in the geotextile tube cross-sectional
area, AT = 9.0 m3.
6.2 Sand Tightness Requirement
Table 2 shows the recommended design criteria for sand tightness requirement (Bezuijen and Vastenburg
2013), which appears to have been derived from the original Pilarczyk (2000) criteria (see Table 1).
Table 2: Recommended design criteria for sand tightness requirement
Hydraulic load Requirement 1 Requirement 2
Stationary load (current) ܱଽ଴ ൏ ͷ‫ܦ‬ଵ଴ ඥ‫ܥ‬௨ ܱଽ଴ ൏ ʹ‫ܦ‬ଽ଴
Dynamic load (wave) ܱଽ଴ ൏ ͳǤͷ‫ܦ‬ଵ଴ ඥ‫ܥ‬௨ ܱଽ଴ ൏ ‫ܦ‬ଽ଴

O90 pore size of the sand filled mattress;


D10 sieve size through which 10% fraction of the sand material passes;
D60 sieve size through which 60% fraction of the sand material passes;
D90 sieve size through which 90% fraction of the sand material passes;
Cu uniformity coefficient (=D60 / D10)
6.3 Hydraulic Stability Requirement Under Wave Attack
The stability of revetment under wave attack is commonly shown in terms of the ratio of wave height to
the weight of the structural units and is a function of the breaker parameter given in the following form:
ுೞ
ൌ ݂ሺ[ሻ [10]
ο೟ ஽ೖ

U௙
ο௧ ൌ െ ͳ [11]
U௪

–ƒ D
[ൌ [12]
ටʹS‫ܪ‬௦ Ȁ‰ܶ௣ ଶ

where, Hs = significant wave height (in m); Dk = effective thickness of the sand-filled geotextile
containment unit (in m); ξ = breaker parameter (-); ∆t = relative buoyant density of the sand-filled
geotextile containment unit (-); ρf = bulk density of sand-fill in geotextile containment unit (in kg/m3); ρw
= density of water (in kg/m3); α = slope angle of the stacked sand-filled geotextile containment units (in
degree); g = gravitational acceleration (in m/s2); and Tp = peak wave period (in s).
6.3.1 Geotextile Bag Revetment Subject to Wave Attack
The stability criteria of geobag revetment under wave attack are given as follows (Reico and Oumeraci
2007, Bezuijen and Vastenburg 2013):
For irregular waves, slope 1:3 or less,
‫ܪ‬௦
൑ ͳǤͶ [13]
ο௧ ‫ܦ‬௞

For slope 1:3 or steeper,
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 229
Protection in Asia
‫ܪ‬௦ ʹǤ͹ͷ
൑ [14]
ο௧ ‫ܦ‬௞ ඥ[

For zone from Hs below SWL to Hs above SWL (see Figure 15),
ுೞ ோ೎
൑ ͲǤ͹ͻ ൅ ͲǤͲͻ [15]
ο೟ ஽ೖ ுೞ

For geotextile bag placement method 1 (see Figure 16a),
Dk = Le sin α [16]
For geotextile bag placement method 2 (see Figure 16b),
Dk = De [17]
where, Hs = significant wave height (in m); Dk = effective thickness of the sand-filled geotextile bag
revetment structure (in m); ∆t = relative buoyant density of the sand-filled geotextile mattress unit (-);
RcLn = distance between crest and still water line (in m); LeLn = effective length of sand-filled geotextile
bag unit (in m); and De Dn = effective thickness of sand-filled geotextile bag unit (in m).

Fig. 15: Applicability zones of stability criteria for geotextile bag revetment

(a) (b)
Fig. 16: Geotextile bag placement (a) method 1 (b) method 2
KN - 230 T.W. Yee

6.3.2 Geotextile Mattress Revetment Subject to Wave Attack


The stability criterion of mattress revetment under wave attack is given as follows (Bezuijen and
Vastenburg 2013):
‫ܪ‬௦ ܵ௠

ο௧ ‫ܦ‬௞ [ଶȀଷ [18]

where, Hs = significant wave height (in m); Dk = effective thickness of the sand-filled geotextile mattress
unit (in m); ∆t = relative buoyant density of the sand-filled geotextile mattress unit (-); and SmLn = stability
factor for the sand-filled geotextile mattress unit (-). The value of SmLn ranges from 4 to 5, depending on
the quality of the sand-fill.
6.3.3 Geotextile Tube Structure Subject to Wave Attack
For stability of sand-filled geotextile tubes under wave attack Bezuijen and Vastenburg (2013)
recommended the use of the following:
‫ܦ‬௘ ܸ ‫ܬ‬௚ U௙ െ U௪
ܶ௚ ൒ ඨʹ U௙ ‰ [19]
ܾܵ ‫ܥ‬ௗ U௪

For sand-filled geotextile tube placed perpendicular to the direction of the wave attack,
Dk = BT [20]
For sand-filled geotextile tube placed parallel to the direction of the wave attack and the length of
geotextile tube is larger than twice the width of the geotextile tube,
Dk = 2BT [21]
For sand-filled geotextile tube placed parallel to the direction of the wave attack and the length of
geotextile tube is smaller than twice the width of the geotextile tube,
Dk = LT [22]
where, Hs = significant wave height (in m); ∆t = relative buoyant density of the sand-filled geotextile
containment unit (-); Dk = effective thickness of revetment (in m); BT = width of the geotextile tube (in
m); and LT = length of the geotextile tube (in m).
6.4 Hydraulic Stability Requirement Under Current Attack
6.4.1 Geotextile Bag and Mattress Revetments Subject to Longitudinal Flow Current Attack
The Pilarczyk relationship can be used to determine the stability of sand-filled geotextile bags and
mattresses subject to longitudinal flow attack and is given as follows (Pilarczyk 2000; Bezuijen and
Vastenburg 2013):
ߔ‫ܭ ்ܭ‬௛ሺ௨೎ೝ ሻమ
ο௧ ‫ܦ‬௞ ൒ ͲǤͲ͵ͷ [23]
ߖ‫ܭ‬௦ ʹ‰
where, ∆t = relative buoyant density of the sand-filled geotextile containment unit (-); Dk = effective
thickness of revetment (in m); Φ = stability parameter (-); Ψ = Shields parameter (-); KT = turbulence
factor (-); Kh = depth parameter (-); Ks = slope parameter (-); ucr = critical flow velocity (in m/s); and g =
gravitational acceleration (in m/s2).
Table 3 shows the values of Φ to adopt in design. Table 4 shows the values of Ψ to adopt in design. Table
5 shows the values of KT to adopt in design.
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 231
Protection in Asia
Table 3: Stability parameter
Revetment location Stability parameter, Φ
For continuous top layer 1.0
For edges 1.5
Table 4: Shields parameter
Revetment type Shields parameter, Ψ
Geotextile bags (< 0.3 m3) 0.035
Geotextile bags (> 0.3 m3) 0.05
Geotextile mattresses 0.07
Table 5: Turbulence factor
Hydraulic situation Turbulence factor, KT
Normal turbulence in rivers 1.0
Higher turbulence: river bends 1.5
Strong turbulence: hydraulic jumps, sharp bends, local disruptions 2.0
Turbulence as the result of propeller jets and other water jets 3.0 to 4.0
By introducing the depth factor, Kh, the depth-averaged flow velocity is translated into a flow velocity just
above the revetment. The depth factor, Kh, is given as follows (Pilarczyk 2000; Bezuijen and Vastenburg
2013):
For developed current profile,
ʹ
‫ܭ‬௛ ൌ  ଶ
ͳʹ݄ [24]
ቆŽ‘‰ ቀ ቁቇ
݇௥

For undeveloped current profile,
݄ ି଴Ǥଶ
‫ܭ‬௛ ൌ ൬ ൰ [25]
݇௥

For a very rough current,
Kh = 1.0 [26]
where, h = water depth (in m); and kr = equivalent roughness according to Nikuradse (in m).
The value of kr depends on the type of revetment. For geotextile bag revetment, the value of kr can be
taken as equal to Dn. For mattress revetment, the value of kr can be taken as equal to De. The slope factor,
Ks, is a function of the influence of the angle of shearing resistance between the structural unit and the
subsoil, and is given as follows (Pilarczyk 2000; Bezuijen and Vastenburg 2013):
For units not anchored at the top of the slope,

•‹ D ଶ
‫ܭ‬௦ ൌ ඨͳ െ ൬ ൰ [27]
•‹ G

For mattresses anchored at the top of the slope and geotextile bags designed with tail reinforcements,
Ks = 1.0 [28]
where, α = slope angle of revetment (in degree); and δ = friction angle between structural unit and the
subsoil (in degree).
KN - 232 T.W. Yee

6.4.2 Geotextile Bags and Tubes Subject to Overtopping Currents


The stability of sand-filled geotextile bags and tubes subject to overtopping currents is commonly shown
in terms of the ratio of critical velocity to the relevant stability factor of the geotextile containment unit
and is represented in the following form:
ܷ௖௥
൒‫ ܨ‬ [29]
ඥ‰ο௧ ‫ܦ‬௞

where, ucr = critical overtopping velocity (in m/s); g = gravitational acceleration (in m/s2); ∆t = relative
buoyant density of the sand-filled geotextile containment unit (-); Dk = effective thickness of the sand-
filled geotextile containment unit (in m); and F = stability factor for the sand-filled geotextile containment
unit (-). Bezuijen and Vastenburg (2013) recommended the use of F = 0.9 for overtopping currents over
sand-filled geotextile bags. Pilarczyk (2000) recommended the use of F = 1.2 for overtopping currents
over sand-filled geotextile tubes. Typically geotextile bags are mechanically filled while geotextile tubes
are hydraulically filled. The higher value of F allowed for geotextile tubes is a reflection of better sand
packing effect when filled hydraulically.
6.5 Geotechnical Stability Requirements
To draw an analogy from design principles for external stability of reinforced soil retaining structures,
the geotextile containment unit be treated as a coherent structure subject to external disturbance (see
Figure 17). For an individual geotextile tube unit an equivalent rectangular section having the same
height, contact width and weight as the analysed geotextile tube geometry may be adopted to simplify
the analysis for checks against sliding, bearing capacity, overturning and global failures. The same
checks can be extended to stacked mounds of geotextile tubes or bags. For the global stability check slip
circle analysis is conducted using standard limit equilibrium design software to assess the stability of a
section geometrically formed using geotextile containment systems. The checks should involve internal,
external and compound failure mechanisms (see Figure 18). Non circular slip planes along the contact
interfaces of geotextile containment units may need checking for as well (see Figure 18). When the slip
planes intersect the geotextile containment units (as may be the case in internal and compound failure
mechanisms) the intersected fabrics may be treated as tensile elements.

(a) (b)

(c) (d)
Fig. 17: Geotechnical stability checks (a) sliding (b) bearing capacity (c) overturning (d) global (Yee 2002)
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 233
Protection in Asia

(a) (b)

(c) (d)
Fig. 18: Global stability failure mechanisms (a) internal (b) compound involving shearing over interfaces
between geotextile containment units (c) compound involving shearing through geotextile containment units
(d) external

6.6 Practical Detailing Requirement


6.6.1 Toe Protection
Toe protection may provide some sliding resistance to the revetment and help to prevent failure of the
revetment from scour (Yee 2012b). Typical forms of toe protection are:
(a) buried toe – the revetment cover layer is extended beneath the bed level to beyond the predicted scour
depth;
(b) sheet piling – this may be used alone or with a concrete toe beam;
(c) toe retaining structure – this may be in the form of a gabion toe, precast concrete mini-wall or
geotextile tube dyke unit;
(d) extension of cover layer along bed – the cover layer may be extended along the bed in front of the
structure; if scour occurs, then this extra length of revetment will drop into the scour hole providing
protection (this should not be used where severe scour may occur as some undermining or loss of
sub-layer material may occur).
6.6.2 Trenching of Geotextile Mattress
The top of sand-filled geotextile mattresses are generally anchored in a trench. The trenching and
anchoring details may differ from case to case, depending on the product, the revetment geometry and
the anchoring resistance force required. When there is a possibility of significant overtopping or overflow
of the structure, erosion protection of the back face may need to be provided for.
KN - 234 T.W. Yee

6.6.3 Termination
Appropriate termination details should be adopted at the ends of the structure to prevent undermining
at the end of the structure. Such details are usually very much site-specific. The termination details may
involve butting against a solid structure; e.g. concrete walls, wing walls of culvert openings, concrete
discharge drains along the bank slopes, etc. or involve the revetment curving back in plan into the bank
slopes.
7. DURABILITY ISSUES OF GEOTEXTILE FILTERS AND CONTAINMENT SYSTEMS
The durability requirement of geotextile containment systems can be most frequently ignored requirement
in design. There are a few reasons for this. Firstly, engineers tend to place more emphasis on quantitative
analysis over qualitative processes. Secondly, this is also an area of knowledge that is generally lacking in
practice (and therefore is discussed in this section as issues rather). Thirdly, the durability requirement of
geotextile containment systems is not universal but very much dependent on project specifics e.g. design
life, length of exposure time, environment, etc. Lastly, any information that exists for specific products
may be closely guarded for commercial reasons because such information usually involves long term
testing and significant time and cost to acquire. The most commonly used polymer for the manufacture of
geotextile filters and containment systems include polypropylene (PP), polyethylene (PE) and polyester
(PET). Therefore discussions on polymer durability will be confined to those concerning PP, PE and
PET.
7.1 Under Covered Conditions
Geotextile filters in riverbanks and coastal applications have a successful track record of about 50 years
worldwide and 40 years in Asia. Except for a relatively short period of exposure during installation, they
remain in a buried or covered condition over their service lives. During the initial years durability of
geotextiles under buried or covered conditions was a concern and major topic of discussion. Today, with
the proven track record the concern has mostly disappeared and the topic of discussion is mostly focussed
on adoption of relevant tests for specification works. This is important because such tests can differentiate
between a geotextile filter with good thermo-oxidative resistance against an inferior one. Additives are
used to fortify polymers to improve specific properties e.g. use of antioxidant stabilisers to improve the
thermo-oxidative resistance of polyolefins. One such test was specifically developed for geotextiles in
marine works. For the Delta works project in the Netherlands in the early 1970s geotextile filters with
a lifetime expectancy of 200 years was sought for. At that time there were no harmonised standards
available for that purpose. Rijkwaterstaat developed the test standard NEN 5132 for that purpose. NEN
5132 allows for differentiation of two different PP yarns for geotextile filters. Those classified as Type A
are projected to have a normal service life expectancy of 30 to 50 years while those classified as Type B
are projected to have a long service life expectancy of 200 to 300 years (Van Santvoort 1994).
7.2 Under Exposed Conditions
The duration of exposure is very dependent on the application. Geotextile filters are exposed for a
relatively short period of time during pre-installation preparation and installation works prior to covering
with armour layers. Geotextile containment systems are relatively exposed for a longer period of time.
For example when geotextile tubes are used as reclamation dykes, typically they may be exposed to 3 to
5 years before they are covered over with the permanent armour layers. For riverbank protection, many
of the geotextile containment systems allow for vegetation growth that would provide cover in the long
term and the period of exposure may be 1 to 2 years, depending on how rapidly the vegetation grows and
matures. However, vegetation does not grow below the normal flow level and during extreme conditions,
water levels do drop below the normal flow levels and certain parts of the geotextile containment structure
may be exposed intermittently. Therefore, the fabric specified for the design of geotextile containment
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 235
Protection in Asia
systems need to incorporate in exposure periods reflecting the project exposure expectations ranging
from 1 to 5 years. More recently, composite fabrics designed specifically for extended exposures of sand-
filled geotextile containment systems have been developed (Hornsey et al 2011). These newly developed
sand-filled geotextile containment systems have allowed designs of structures with a few decades of
exposure.
7.2.1 UV Resistance
The solar radiation that is intercepted by the Earth’s thermopause is known as the thermopause insolation
(intercepted solar radiation) and comprises of 8% combined gamma-rays, X-rays & UV (ultraviolet)
radiation, 47% visible light and 45% infrared wavelength radiation (Christopherson 2000). Gamma-rays
and X-rays have wavelengths smaller than 100 nm. UV radiation wavelengths range from 100 to 400 nm
and is subdivided into UV-A (315 to 400 nm), UV-B (280 to 315 nm) and UV-C (100 to 280 nm). Visible
light has wavelength ranging from 400 to 700 nm while infrared wavelengths are larger than 700 nm. The
solar radiation that reaches the Earth’s surface has a different composition and intensity from that of the
thermopause insolation as radiation of certain wavelengths is absorbed by the Earth’s atmosphere. The
radiation can also be regarded as individual photons whose energies are highest at the shorter wavelengths
and lowest for the longer ones. The extremely harmful gamma-ray and X-ray are absorbed by the upper
half of the atmosphere while the most harmful UV component, UV-C is absorbed by the ozone layer of
the atmosphere. Infrared radiation of certain wavelengths also reaches the Earth’s surface at reduced
intensity due to absorption by the atmosphere. The amount of radiation and the radiation spectrum that
reaches the Earth’s surface is dependent on factors like the Sun’s elevation, the surface location latitude
and altitude, cloud cover, etc. Figure 19 shows the map of Earth’s surface annual average insolation in
kWh/m2/day over 22 years.

Fig. 19: Map of Earth’s surface Earth’s surface annual average insolation in kWh/m2/day over 22 years
(Chandler et al 2010)
The UV radiation that reaches Earth’s surface is of concern because it causes bond breakage in the
polymer used in the manufacture of geotextile filters and containment systems leading to loss of all
properties including discolouration, tensile strength and tensile elongation. However different polymer
resins have different degeneration mechanisms and wavelength sensitivities. PE degenerates through
the photo-oxidation process which is initiated by the generation of free radicals that are responsible for
starting a chain oxidation reaction leading to breakage of the polymer backbone (CUR 243, 2012). PE
is most sensitive to wavelengths in the range of 330 to 360 nm. PP also degenerates through the photo-
oxidation process but has a narrower wavelength sensitivity range of 355 to 360 nm. PET degenerates
KN - 236 T.W. Yee

through polymer chain scission leading to reduction in strength and the generation of carboxyl end groups
and is most sensitive around 325 nm.
Outdoor natural weathering is commonly used to test polymer and paints as it is easy to set up, cheap to
run and represents reality. The main disadvantage is that the time consumed is long to get meaningful
results and it is difficult to factor in variability in geography and weather changes. For example, 3 months
of exposure in winter is not equivalent to 3 months of exposure in summer and 3 months of exposure
in Hokkaido is not the same as 3 months in Kalimantan. Accelerated UV exposure tests are commonly
used to assess the UV degradation property of geotextiles. There are two principal types of accelerated
weathering equipment i.e. Xenon arc equipment and fluorescent UV equipment. The test standards most
commonly adopted for testing UV degradation of geotextiles are ASTM D4355 which uses the Xenon
arc equipment and EN 12224 which uses the fluorescent UV equipment. The ASTM D4355 test method
using simulated UV radiation with irradiance of 0.35 W/m2 at 340 nm has a test cycle of 500 hours and
translates to 52 MJ/m2 of UV radiant exposure. The EN 12224 test method using UVA-340 lamp simulated
UV radiation has a test cycle of 430 hours and translates to 50 MJ/m2 of UV radiant exposure. The results
are presented as a residual percentage in terms of tensile strength at the end of the test cycle. These tests
simulate an equivalent outdoor natural weathering of about 2 to 6 months, depending on various factors
mentioned earlier that determine the amount of radiation reaches the Earth’s surface. Geotextile filters are
typically subject to a relatively short exposure time of a few weeks and therefore such test methods are very
relevant for the setting material specification. These test methods are also useful for making comparisons
between various geotextiles. For geotextile containment systems exposure time may involve a few years
to even decades depending on the design. The residual tensile strength derived for a single test cycle
may not be very meaningful for the assessment of UV degradation resistance of a geotextile containment
product or its suitability for a specific project design. Extended tests involving multiple cycles may be
required for the proper assessment of UV degradation resistance of a geotextile containment product and
its suitability for a specific project. Figure 20 shows the extended UV degradation tests results (EN 12224
test method) for a geotextile filter and two geotextile tube products. The UV resistance performance and
estimated half lives of the three products are given in Table 6. The annual solar insolation at a project site
may be estimated using Figure 19. The conversion of units are as follows; 1 kW/m2/day = 365 kWh/m2/
year = 3.6 MJ/m2/year. The ratio of UV radiation to total solar radiation for South Florida was determined
as 1:20 in the studies by Baker (1997). The same ratio is assumed to apply to the project sites presented
in Table 6. Geotextile filter A refers to the product used for project shown in Figure 7a. Geotextile tube
B refers to the product used for Lach Huyen Bridge Project presented in Section 8.2. Geotextile tube C
refers to the product used for project shown in Figure 12a.

Fig. 20: Extended UV degradation tests using EN 12224 test method for
a geotextile filter and two geotextile tube products
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 237
Protection in Asia
Table 6: UV resistance performance and estimated half lives of three products
Product Polymer Fabric type Project Annual solar Annual Time to Projected
base site insolation at site UV 90 % UV half
radiation strength life at site
at site at site

(kWh/ (MJ/m2/ (MJ/m2/ (year) (year)


m2/day) year) year)
Geotextile PP with Nonwoven Singapore 5 6570 328.5 0.06 0.3
filter A moderate
UV additive
Geotextile PP with Woven Haiphong, 4.5 5913 295.7 1.7 8.5
tube B heavy UV Vietnam
additive
Geotextile PP with Woven + Gan, Addu 6 7884 394.2 > 10 > 50
tube C heavy UV coarse fibre Atoll, (projected)
additive composite Maldives
7.2.2 Abrasion Resistance
Abrasion due to water borne sand, shell and coral fragments is a threat unique to sand-filled geotextile
containment systems (Hornsey et al 2011). The abrasion test (commonly known as the German rotating
drum test) BAW RPG 3.11 according to Federal Waterways Engineering and Research Institute–
Guidelines for Testing Geotextiles for Navigable Waterways is commonly used (BAW 1994). BAW
RPG 3.11 was specifically developed for assessing abrasion resistance of geotextiles used in coastal and
riverbank applications. The test method simulates abrasion loads on geotextiles such as that caused by
the movement of rocks in a revetment. A mixture of stone chippings and water passes over geotextile
samples installed in a rotation drum for 40,000 revolutions in an anti-clockwise direction and repeated
for another 40,000 revolutions in a clockwise direction. The abrasion resistance is measured in terms
of the percentage ratio of the after-test sample tensile strength to the virgin before-test sample tensile
strength. Hornsey et al (2011) recommended a minimum strength retention value of 70% for sand-filled
geotextile containment systems under exposed coastal applications. The fabric used for geotextile tube
B in the Vietnam project (see Table 6) recorded a strength retention value of 76% while the fabric used
for geotextile tube C in the Maldives project (see Table 6) recorded a strength retention value of 99% in
laboratory tests.
7.2.3 Damage Resistance Against Incidental Impacts and Vandalism
Currently, no test standards that can be used to index against mechanical damage during exposure exist.
Mechanical damage may be caused by many means e.g. airborne and waterborne impacting debris during
severe storm events, vessel impact, human mischief actions, etc. During the last decade composite fabrics
that consist of a standard inner fabric with a coarse fibre outer layer were developed e.g. geotextile tube C in
the Maldives project (see Table 6). In its pristine form the composite has generally improved performance
against mechanical damage. In addition the outer coarse fibre layer allows sand to be trapped on the
geotextile containment fabric that further enhances not only the damage resistance against incidental
impacts but also protection against vandal knife cuts (Hornsey et al 2011, Saathoff et al 2007). There have
been attempts to use existing test standards with some modifications to index against mechanical damage.
For example Hornsey et al (2011) proposed to combine CBR strength and elongation to indicate impact
energy. However, a static force is used for the CBR test and therefore may not reflect the dynamic force
of impacting debris during severe storm events.
KN - 238 T.W. Yee

The Southeast coast of United States is prone to hurricanes. Conventional wisdom used to be that winds
themselves were responsible for most damage until Hurricane Andrew’s devastation of south Florida
in 1992 changed the thinking. Post-disaster investigations indicated that much of the damage occurred
because windows and doors were compromised by wind-borne debris. ASTM E1996-03 Standard
Specification for Performance of Exterior Windows, Curtain Walls, Doors and Storm Shutters Impacted
by Windborne Debris in Hurricanes was developed for the purpose of assessing wind-borne debris on
shutters, doors and windows. This test is currently used as the standard test to assess the performance of
porous impact protective systems to shutters, doors and windows. Recent developmental work involves
adapting ASTM E1996-03 to assess impact damage resistance of sand-filled geotextile containment
systems. Figure 21a shows a timber missile just before impact of a sand-filled geotextile bag sample made
from the same fabric used for geotextile tube C in the Maldives project (see Table 6). Figure 21b shows
the close up of the sand-filled geotextile sample after impact. The ASTM E1996-03 classifies 5 different
missile levels. The highest is missile level E which specifies a 4.1 kg, 50 mm x 100 mm x 2.4 m timber
section fired from a cannon that gives a missile impact speed of 24.4 m/s; recommended for enhanced
protection of essential facilities in conditions associated up to Wind Zone 4 (basic wind speed of 140 mph
or 63 m/s). The sample representing geotextile tube C was tested at missile impact speed in excess of 30
m/s and survived without damage.

(a) (b)
Fig. 21: Timber missile test (a) showing timber missile just before impact (b) showing close up after impact

8. CASE STUDIES
8.1 Kerteh Bay Breakwater Project, Malaysia
The Kerteh bay breakwater project has been reported by Tilmans et al (1992) and Pilarczyk and Zeidler
(1996) previously.
8.1.1 Project Brief
The Rantau Petronas Complex, which consists of housing facilities, a school complex and a golf course,
is situated within Kerteh Bay on the east coast of Peninsular Malaysia. Much of the coast consists of a
series of large and small hook-shaped bays. In recent decades Kerteh Bay has suffered from accelerated
severe coastal erosion. The bay is exposed to direct wave attack. During the annual northeast monsoons,
storms blowing from the South China Sea were of such severity that the sea would encroach the beach up
to the very limit of Rantau Petronas Complex.
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 239
Protection in Asia
8.1.2 Site Conditions
Kerteh Bay represents a geomorphorlogy typical of coasts whose development is controlled by protruding
headlands. The stretch of beach to be restored and preserved is approximately 2.1 km. The site is generally
blanketed by coarse to fine sand. The typical cross sectional bathymetry and the hydraulic conditions are
as shown in Figure 22. The water level fluctuations are mainly caused by astronomical tides, which
are of the mixed type; diurnal during spring and semi-diurnal during neap. However, storm surge has a
significant effect during the northeast monsoon season. The highest astronomical tide was predicted at
2.7 m + ACD. Coupled with a 40 years return period design surge of 0.8 m, the design water level was
3.5 m + ACD.

Fig. 22: Cross-section and hydraulic conditions at Kerteh Bay, Malaysia


Predominant waves approach from the directions of 30ON and 60ON. Controlling wave climate was
derived from approximately 12,000 ship observations between 1960 and 1990 offshore of the area of
study. The data was transformed to Kerteh Bay, taking into account shoaling, refraction and dissipation.
The extreme design waves for 30 years return period impacting the area was determined as follows:
- from the direction of 60ON, significant wave height, Hs = 4.0 m, return period, T = 9 s
- from the direction of 30ON, significant wave height, Hs = 6.4 m, return period, T = 10 s
The role of tidal currents was found to be minor compared to storm waves.
For the dynamically stable configuration of Kerteh Bay, the littoral drift rate was assessed at about 210,000
m3 per year, of which 80% happens during the northeast monsoon season. The up-coast sediment supply
is from Paka Bay via offshore bar bypassing at the northern end of Kerteh Bay. Erosion observations
indicated an average up-coast sediment supply deficit of about 40,000 m3 per year to Kerteh Bay. The
cause and persistency of the deficit was deemed likely to be due to shore developments within the up-
coast Paka Bay.
8.1.3 Coastal Erosion Protection Options
There were four options evaluated for the restoration and preservation of the coastline in front of Rantau
Petronas Complex. Large scale beach nourishment was deemed to provide good beach aesthetics with
KN - 240 T.W. Yee

minimal overall environmental impact but would be costly over the long run as the exercise would need
to be repeated in the future on a regular basis as the sea would reclaim the nourished beach over time.
Coastal revetment would protect the complex but would result in poor beach aesthetics. Also, the coastal
erosion problem would simply be shifted down-coast. A groyne (perpendicularly or sub-perpendicularly
shore attached dyke) system would incur low initial construction cost but provide poor beach aesthetics
and the coastal erosion problem would also simply be shifted down-coast. An offshore breakwater system
may incur high initial cost but was chosen because the solution would result in lowest overall cost (when
long term maintenance cost are accounted for) with minimal environmental impact as well as providing
good beach aesthetics.
8.1.4 Construction
The resulting design called for the construction of three detached offshore breakwaters (see Figure 23).
Initial nourishments of 100,000 m3 of sand was placed at the north side, 200,000 m3 of sand was placed
at the south side while a further 100,000 m3 of sand was spread along the coastline. Figure 24 shows the
cross-section of one of the breakwaters constructed. The geotextile used was a woven polypropylene
geotextile with warp tensile strength of 200 kN/m and weft tensile strength of 40 kN/m. The geotextile
had in-woven loops to allow the bamboo to be tied to the geotextile at 0.5 m grid centers (see Figure 25a).
The basal geotextile was first made into a 30 m wide fascine mattress by tying bamboo to the geotextile
onshore (see Figure 25b). The geotextile/bamboo fascine mattress was first floated out to sea (see Figure
25c) and installed by rock ballasting into position (see Figure 25d). The project involved the use of
70,000 m2 geotextile. A total of 160,000 tons of quarried rocks was used. The volume of material dredged
was 500,000 m3 and 400,000 m3 of sand was used for initial beach nourishment to check the acute coastal
erosion at Rantau Petronas Complex.

Fig. 23: Layout of offshore breakwaters at Kerteh Bay, Malaysia

Fig. 24: Cross-section of breakwater at Kerteh Bay, Malaysia


Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 241
Protection in Asia

(a) (b)

(c) (d)
Fig. 25: Installation of basal geotextile filter for Kerteh breakwater (a) detail of in-woven fabric loop and
rope attachment (b) tying bamboo to looped geotextile filter (c) launching prepared fascine mattress basal
geotextile filter into the sea (d) bamboo fascine mattress basal geotextile filter to be sunk onto the seabed
using rock ballast

8.2 Lach Huyen Bridge Project, Vietnam


The Lach Huyen bridge project has been reported by Yee (2016) previously.
8.2.1 Project Brief
Hai Phong Port and Cai Lan Port are the two ports currently serving the northern region of Vietnam. It is
estimated that demand for containerized cargo in the northern region of Vietnam will be 42 million tons
in 2015, which will further increase to 59 million tons by 2020 in tandem with economic growth, well
beyond the combined capacities of the two existing ports. Additionally, the growing trend in international
sea transport markets in recent years is that shipping companies have been increasing their orders for
large-container vessels as their measure to meet customers’ needs and also from the viewpoint of cost
reductions. In order to upgrade the functions of the northern region of Vietnam as one of the international
distribution centers, it is necessary to develop a port that has sufficient depth to accommodate large
container vessels. The new Lach Huyen Port located on the eastern side of Cat Hai Island was planned
and designed as a deep sea port to handle larger vessels and to cope with the projected growth of demand
in cargo volume.
Package 6 of the overall Lach Huyen Port Infrastructure Construction Project, involves the construction
of a new 15.6 km long highway linking the Hanoi-Haiphong Expressway at the Tan Vu Interchange on
the mainland to Lach Huyen Port. The proposed Tan Vu-Lach Huyen Highway provides a shortened link
KN - 242 T.W. Yee

between the Hai An District and Cat Hai Island. The travel time between Tan Vu Interchange and Lach
Huyen Port will be reduced from the 2011 estimate of 2.5 hours to about 15 minutes with the completion
of the new highway. Figure 26 shows the layout of the Tan Vu-Lach Huyen Highway. The highway
includes the 5.4 km long Lach Huyen Bridge with 84 spans supported by piers. The bridge site spans
across the combined estuary of Bach Dang River and Cam River. The Lach Huyen Bridge consists of the
4.4 km western approach bridge, the 0.5 km main bridge (over the navigational channel) and the 0.5 km
eastern approach bridge. The western abutment of Lach Huyen Bridge is located just south of Dinh Vu
Development Area while the eastern abutment is located at the western shore of Cat Hai Island. Upon
completion in 2017, Lach Huyen Bridge will be the longest sea-crossing bridge in Vietnam. The bridge
will be 16 m wide accommodating four lanes for vehicles and two safety corridors while the approaches
will be 29.5 m wide for traffic design speed of 80 kph.

Fig. 26: Layout of Tan Vu-Lach Huyen Highway


8.2.2 Site Conditions
The area consists of Quaternary coastal sediments overlying a Pleistocene land surface, and represents
the Holocene marine transgression and regression. Up until 6000 years ago, the sea rose to around or
above its present elevation, converting the Pleistocene terrestrial landscape to a Holocene tidal landscape
of tidal flat, channel and mangrove environments. Sea level lowering at around 4000 years ago triggered a
switch in the dominant sedimentary processes, allowing floodplain sediments to be deposited increasingly
seawards.
Haiphong has a tropical climate with high humidity and temperatures. The average annual rainfall is
1760 mm (over 76 years of observations). The rainy season covers the months from May to October and
accounts for about 80% of the total annual rainfall. The dry season is from November to March. The
months from October to December are particularly foggy. For about twenty days of the year the visibility
is less than one km, occurring mostly during October to December. Generally, the winds in the Haiphong
area are gentle. Winds of 1 to 4 m/s account for 60% of occurrences while winds over 10 m/s account for
only 2% of occurrences. According to the wind records in Hon Dau Observatory the dominant directions
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 243
Protection in Asia
are East, Southeast, South and North. The area is occasionally subject to typhoons, typically from June to
September. The strongest measured winds induced by typhoons reached 51 m/s on August 21, 1977.
Haiphong has the diurnal tidal regime i.e. one high tide and one low tide every day. The spring tide/neap
tide cycle is 14 days with a maximum tidal range of nearly 4 m. The highest high water level (HHWL) is
at +2.55 m. The mean high water level (MHWL) is at +1.97 m. The mean low water level (MLWL) is at
-1.67 m .The mean tide level (MTL) is at +0.15 m. The Hon Dau Station records of 2006 to 2008 show
that 60% of waves come from directions of the quadrant of East to South. Wave heights of more than
1 m occur about 9% of the time. The significant wave height, Hs, is 1.7 m with a peak wave period of
11.16 s. The tidal wave propagates from South to North with mean velocities between 0.2 to 0.3 m/s. The
maximum ebb-tidal current is 0.6 m/s while the maximum flood-tidal current is 0.5 m/s.
The subsoil in the area generally consists of very thick alluvial and marine clay deposits above a dense
to very dense sand/gravel foundation. Generally, there are 4 distinct layers of the deposits. The first layer
consists of very soft organic sandy silty clay with thickness of about 6 to 9 m and undrained shear strength
of about 5 kN/m2. The second layer consists of soft to very soft clay with fine sand with thickness of
about 6 to 8 m and undrained shear strength of about 10 kN/m2 (Montulet et al 2013). Below that is an
approximately 4.5 m layer of stiff to very stiff clay with gravel. The fourth layer generally consists of
medium stiff clay.
8.2.3 Bridge Construction Challenges
Bridge construction works require heavy machinery for installation of foundation piles, construction
of bridge piers and launching of bridge sections. Soft ground and tidal conditions are the two main
construction challenges imposed by the site environment. The soft ground conditions create difficulty for
machinery to work on. A significant portion of the bridge alignment is under shallow water either part of
the day or all the time. Except for the navigation channel under the main bridge, there is insufficient water
depth for access for work barges during all tidal conditions. Both land-based and offshore construction
methods are employed along different segments of the bridge alignment. Figure 27 shows the adoption of
land-based construction and offshore construction methods for Lach Huyen Bridge.
To allow bridge construction works to be carried out on dry land expediently two construction works
platforms with approximate width of 26 m and 3.9 km in total length were reclaimed along the alignment


Fig. 27: Layout of Lach Huyen Bridge and location indicating adoption of land-based construction and
offshore construction methods
KN - 244 T.W. Yee

of the bridge. A portion of 1.5 km of the bridge was constructed by the offshore method using work
barges. A 1 km long channel along the alignment of the bridge was dredged to increase water depth
sufficiently to allow the work barges to operate without any low tide interruptions. The dredging works
of the 1 km access channel were carried out using front arm excavators placed on a flat bottom barge. The
dredged sediment is pumped to the dredged sediment containment facility.
8.2.4 Construction of Works Platform and Dredged Sediment Containment Facility
Figure 28 shows the typical cross section of the construction works platform constructed using geotextile
tubes as reclamation bunds. This construction technique is used for the first time in Vietnam but has
been used widely in Korea in the past (Yee et al, 2007; Yee & Choi, 2008). The geotextile tubes were
hydraulically filled with sand dredged offshore. About 27 km of geotextile tubes were used as reclamation
bund, segmentally stacked up to 5 layers high over intertidal soft clay deposits for the construction
of the platforms. The geotextile tubes used for the construction of the reclamation bund comprised of
circumferences of 4.6, 6, 7.5 and 9.5 m, with typical lengths 50 m. The works platform together with the
geotextile tube reclamation bund will be embedded within a future land reclamation that would enlarge the
Dinh Vu Development Area to approximately double of its current size (see Figure 26). The installation
of geotextile tubes began in July 2014 and was completed by December 2014. Sand is dredged from the
channels nearby. Dredging is carried out using a suction dredger and placed inside a sand supply barge.
When the sand supply barge is full it would sail to the unloading jetty for transfer to the sand stockpile
depot or to a specific work area where the sand is needed. From the stockpile depot, sand is pumped
through about 3 km of HDPE pipeline with a diameter of 300 mm to the temporary storage yard. A
booster pump is used midway to maintain sufficient pressure throughout the pipeline. When required,
sand is pumped from the temporary storage yard for the installation of geotextile tube and reclamation
of works platform (see Figure 27). Figure 29a shows the laying out of geotextile tube for filling with
sand hydraulically. Figure 29b shows the placement of hydraulic fill between installed geotextile tubes.
Figures 29c and 29d show the works platform in progress westward and eastward respectively.

Fig. 28: Typical cross section of the temporary works platform constructed with geotextile tubes as
reclamation bunds
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 245
Protection in Asia

(b) (b)

(c) (d)
Fig. 29: Construction works using geotextile tubes (a) laying out geotextile tube ready for filling (b)
hydraulic fill placed between installed geotextile tubes (c) works platform progress westward (d) works
platform progress eastward

Geotextile tubes were used to construct the perimeter bund of the containment facility. The bund was
constructed using geotextile tube with standard length of 15 m and circumference of 9.5 m. The bottom
level consists of two units placed side by side and a top unit is then placed centrally above that. The
containment facility has a storage capacity of more than 600,000 m3. Figure 30a shows the cross section
of the dredged sediment containment facility. Figure 30b shows the photo of the completed dredged
sediment containment facility. Figure 31 shows the overall view of the reclaimed works platform and the
bridge construction works in an advanced stage; note the dredged sediment containment facility partially
seen at the top left of diagram.

(a) (b)
Fig. 30: Dredged sediment containment facility (a) cross-section (b) completed disposal area
KN - 246 T.W. Yee

Fig. 31: Overall view of the reclaimed works platform and the bridge construction works in progress

9. CONCLUSIONS
Geotextile filters were first introduced to Asia about 40 years ago for riverbank and coastal erosion
applications. These applications cover revetment geotextile filters and dyke and breakwater basal
geotextile filters. Geotextile containment systems followed later and together with geotextile filters,
they have become important solutions to problems concerning riverbank and coastal erosion. Today,
geotextile filters and geotextile containment systems are used for a wide range of erosion protection
applications ranging from simple, low cost applications to sophisticated engineering solutions. Many of
the more recent applications have illustrated the innovative use of these materials. Further innovation will
be anticipated in the future as existing sources of rock become more difficult of obtain and less attractive
from an environmental sustainability point of consideration.
Acknowledgements
The author would like to acknowledge the assistance of the management and construction team of
Sumitomo-Mitsui Construction Co., Ltd. at Lach Huyen Bridge Project for the sharing of information
and thank them for the hospitality during the author’s participation at site. The author would also like to
acknowledge Edwin Zengerink and Chris Timpson of TenCate Geosynthetics for their contribution to the
test results and discussions on durability issues.
REFERENCES
AASHTO M288. Standard Specification for Geotextile Specification for Highway Applications, American
Association of State Highway and Transportation Officials (AASHTO), Washington, DC., USA.
ASTM D 4355. Standard Test Method for Deterioration of Geotextiles by Exposure to Light, Moisture
and Heat in a Xenon Arc Type Apparatus, American Society for Testing and Materials, West
Conshohocken, Pennsylvania, USA.
ASTM D 5101. Standard Test Method for Measuring the Soil-Geotextile System Clogging Potential by
the Gradient Ratio, American Society for Testing and Materials, West Conshohocken, Pennsylvania,
USA.
ASTM D 5567. Standard Test Method for Hydraulic Conductivity Ratio (HCR) Testing of Soil/Geotextile
Systems, American Society for Testing and Materials, West Conshohocken, Pennsylvania, USA.
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 247
Protection in Asia
ASTM E 1996-03. Standard Specification for Performance of Exterior Windows, Curtain Walls, Doors
and Storm Shutters Impacted by Windborne Debris in Hurricanes, American Society for Testing and
Materials, West Conshohocken, Pennsylvania, USA.
Aydilek, A.H. and Edil, T.B. (2002). Filtration Performance of Woven Geotextiles with Wastewater
Treatment Sludge. Journal Geosynthetics International, Vol. 9, No. 1, pp. 41-69.
Baker, T.L. (1997). Long-Term Relationship of Outdoor Exposure to Xenon-Arc Test Apparatus Exposure.
Geosynthetics ’97, Longbeach, California, USA, pp. 177-190.
BAW (1993). Code of Practice – Use of Geotextile Filters on Waterways - MAG. Federal Waterways
Engineering and Research Institute - BAW, Hamburg, Germany.
BAW (1994). Guidelines for Testing Geotextiles for Navigable Waterways - RPG. Federal Waterways
Engineering and Research Institute - BAW, Hamburg, Germany.
Berendsen E. (1996) Dumping of rock on geotextiles. Geosynthetics: Applications, Design and
Construction (ed. De Groot, Den Hoedt & Termaat), Balkema, Rotterdam, The Netherlands, pp.
919-924.
Bezuijen, A., & Vastenburg, E.W. (2013). Geosystems: design rules and applications, CRC/Balkema,
Leiden, The Netherlands.
Bhatia, S.K. and Smith, J.L. (1996). Geotextile Characterization and Pore-Size Distribution: Part I. A
Review of Manufacturing Processes. Journal Geosynthetics International, Vol. 3, No. 1, pp. 85-105.
Cantré, S., & Saathoff, F. (2011). Design method for geotextile tubes considering strain – formulation
and verification by laboratory tests using photogrammetry. Journal Geotextiles and Geomembranes,
29 (3), pp. 201–210.
Carroll, R.G., Jr. (1983). Geotextile Filter Criteria, Transportation Research Record 916, Washington,
DC, USA, pp. 46-53.
Chandler, W.S., Hoell, J.M., Westberg, D., Whitlock, C.H., Zhang, T. and Stackhouse, P.W., Jr. (2010).
Near Real-Time Global Radiation and Meteorology Web Services Available from NASA. American
Solar Energy Society, USA.
Chew, S.H., Karunaratne, G.P., Tan, S.A. and Wong, W.K. (1999). Standardised drop test to evaluate
the puncture resistance of geotextiles in coastal revetments. Recontres Geosynthetics 99, Bordeaux,
France, pp. 303-310.
Christopher, B.R. and Holtz, R.D. (1985). Geotextile Engineering Manual. Report No. FHWA-TS-86/203,
March 1985, U.S. Department of Transportation, Federal Highway Administration, Washington, DC,
USA.
Christopherson, R.W. (2000). Geosystems – An Introduction to Physical Geography. Fourth Edition,
Prentice Hall, New Jersey, USA.
CIRIA (2007). The rock manual, the use of rock in hydraulic engineering 2nd Edition, C683, CIRIA,
UK.
CUR 243 (2012) Durability of geosynthetics, CUR committee C 187, Gouda, The Netherlands.
Delmas, P., Artieres, O., Schorgenhuber, H. and Lugmayr, R. (2000). Development of a new geotextile
filtration system. Geofilters 2000, Warsaw, Poland, pp. 51-58.
EN 12224 (2000). Geotextiles and geotextile-related products – Determination of the resistance to
weathering. European Standards.
KN - 248 T.W. Yee

Fischer, G.R., Christopher, B.R. and Holtz, R.D. (1990). Filter Criteria Based on Pore Size Distribution.
Fourth International Conference on Geotextiles, Geomembranes and Related Products, Balkema,
Vol. 1, The Hague, The Netherlands, pp. 289-94.
Giroud, J.P. (1996). Granular Filters and Geotextile Filters. Geofilters ‘96, Montreal, Canada, pp. 565-
680.
Giroud, J.P., Delmas, P. and Artieres, O. (1998) Theoretical basis for the development of a two-layer
geotextile filter. Sixth International Conference on Geosynthetics, Atlanta, USA, pp. 1037-1044.
GRI (1991). GRI Test Method GT11: Standard practice for installation of geotextile tubes used as coastal
and riverine structures. Geosynthetics Institute, Folsom, Pennsylvania, USA.
Guo, W., Chu, J. & Yan, S.W. (2011). Effect of subgrade stiffness on analysis of geosynthetic tube.
Journal Geotextiles and Geomembrane, 29 (2014), pp. 277-284.
Hardman, E. (2000). Textiles in filtration. Handbook of technical textiles, Chapter 13 (ed. A.R. Horrocks
& S.C. Anand), The Textile Institute, CRC Press, Cambridge, UK, pp. 316-371.
HCZMP (1993). Draft Coastal Hazard Assessment Project. Hawaii Coastal Zone Management Program
(HCZMP), Hawaii, USA.
Heibum, M. (2014). Geosynthetics for waterways and flood protection structures – controlling the
interaction of water and soil. Journal Geotextiles and Geomembrane, 42 (2014), pp. 374-393.
Heibum, M., Fourie, A., Girard, H., Karunaratne, G.P., Lafleur, J. and Palmeira, E.M. (2006). Hydraulic
applications of geosynthetics. Eighth International Conference on Geosynthetics, Yokohama, Japan.
Milpress, Rotterdam, pp. 79-120.
Hornsey, W.P., Carley, J.T., Coghlan, I.R. and Cox, R.J. (2011). Geotextile sand container shoreline
protection systems: Design and application. Journal Geotextiles and Geomembranes, 29 (2011) pp.
425-439.
ISO 9864: 2005. Geosynthetics – test method for determination of mass per unit are of geotextiles and
geotextile-related products. International Standards Organisation.
Kazimierowicz, K. (1994). Simple analysis of deformation of sand sausages. Fifth International
Conference on Geosynthetics, Singapore, Balkema, Vol. 2, pp. 775–778.
Koerner, R.M. and Ko, F.M. (1982). Laboratory Studies on Long-Term Drainage Capability of Geotextiles.
Second International Conference on Geotextiles, IFAI, Vol. 1, Las Vegas, Nevada, USA, pp. 91-95.
Lawson, C.R. (1982). Geotextile requirements for erosion control structures. Recent Developments in
Ground Improvement Techniques, Balkema, Rotterdam, The Netherlands, pp. 177-192.
Lawson, C.R. (1992). Geotextile revetment filters. Journal Geotextiles and Geomembranes, Vol. 11, Nos.
4-6, Elsevier, pp. 431-448.
Lawson, C.R. (2008). Geotextile containment for hydraulic and environmental engineering. Journal
Geosynthetics International, 15, No. 6, pp. 384-427.
Lawson, C.R. (2012). Geosynthetics for riverbank and coastal protection in Asia. Fifth Asian Regional
Conference on Geosynthetics, Bangkok, Thailand, pp. 25-46.
Leschinsky, D., & Leschinsky, O. (1995). Geosynthetic confined pressurized slurry (GeoCoPS):
Supplemental notes for version 1.0. Report TR CPARGL-96-1, U.S. Army Engineer Waterways
Experiment Station, Vicksburg, Mississippi USA.
Geotextile Filters and Containment Systems for Riverbank and Coastal Erosion KN - 249
Protection in Asia
Leschinsky, D., Leschinsky, O., Ling, H. I. & Gilbert, P. A. (1996). Geosynthetic tubes for confining
pressurized slurry: some design aspects. Journal of Geotechnical Engineering Division, American
Society of Civil Engineers, 122, No. 8, pp. 682-690.
Lipidus, D.F. (1990). Collins Dictionary of Geology, Collins, Glasgow, UK.
McConnell, K. (1998). Revetment systems against wave attack – A design manual, HR Wallingford,
Thomas Telford, London, UK.
Montulet, A., Halleux, L., and Nguyen, V.N. (2013). Subsoil improvement works at DVIZ, Haiphong,
Vietnam. Journal Terra et Aqua, IADC, The Netherlands, No. 133, December 2013, pp. 3-9.
NEN 5132 (1992). Geotextiles - Polypropylene yarns for fabrics - Requirements and test methods.
Nederlands Norm. (in Dutch).
Reico, J. and Oumeraci, H. (2007). Effect of deformations on hydraulic stability of coastal structures
made of geotextile sand containers. Journal Geotextiles and Geomembranes, 25, Elsevier, pp. 278-
292.
Palmerton, J. B. (1998). SOFFTWIN - simulation of fluid filled tubes for windows. Computer Program,
Copyright 1998.
Pidwirny, M. (1999). Fundamentals of Physical Geography. Chapter 10: Introduction to the Lithosphere,
(w) Erosion and Deposition, PhysicalGeography.Net.
Pilarczyk, K.W. (2000). Geosynthetics and Geosystems in Hydraulic and Coastal Engineering, A.A.
Balkema, Rotterdam, The Netherlands.
Pilarczyk, K.W. & Zeidler, R.B. (1996). Offshore Breakwaters and Shore Evolution Control, Balkema:
Rotterdam, The Netherlands.
Plaut, R. H. and Suherman, S. (1998). Two-dimensional analysis of geosynthetic tubes. Journal Acta
Mechanica, No. 3–4, pp. 207-218.
Przedwojski, B., Blazejewski, R. and Pilarczyk, K.W. (1995). Specification of materials and elements
of structures, Chapter 7, River Training Techniques – Fundamentals, Design and Applications, A.A.
Balkema, Rotterdam, The Netherlands, pp. 487-516.
Saathoff, F., Oumeraci, H. and Restall, S. (2007) Australian and German experiences on the use of
geotextile containers. Journal Geotextiles and Geomembranes, 25, Elsevier, pp. 251-263.
Tilmans, W.M.K., Klomp, W.H.G. and de Vroeg, H.H. (1992). Coastal erosion – the Kerteh case: Analsyis
of causative factors and mitigative measures using dedicated modelling tools. International Colloqium
on Computer Applications in Coastal and Offshore Engineering, Kuala Lumpur, Malaysia.
Timoshenko, S., & Woinowsky-Krieger, S. (1959). Theory of Plates and Shells, McGraw-Hill, New
York, USA.
Van Santvoort, G.P.T.M. (1994). Geotextiles and Geomembranes in Civil Engineering, A.A. Balkema,
Rotterdam, The Netherlands.
Watn, A. and Chew (2002). Geosynthetic damage – from laboratory to field. Seventh International
Conference on Geosynthetics, Nice, France, Vol. 4, pp. 1203-1226.
Wong, W.K., Chew, S.H., Tan, S.A. and Faure, Y.H. (2002). Puncture resistance of geotextiles against
installation. Seventh International Conference on Geosynthetics, Nice, France, Vol. 4, pp. 1405-
1408.
KN - 250 T.W. Yee

Yee, T.W. (2002). Construction of underwater dykes using geotextile containment systems. Seventh
International Conference on Geosynthetics, Nice, France, Vol. 3, pp. 1161-1164.
Yee, T.W. (2012a). Analysis of geotextile tube using discrete membrane elements method. Fifth European
Geosynthetics Congress, Valencia, Spain, Vol. 5, pp. 583-587.
Yee, T.W. (2012b). Geosynthetics for erosion control in hydraulic environment. Fifth Asian Regional
Conference on Geosynthetics, Bangkok, Thailand, pp. 119-134.
Yee, T.W. (2016). Geotextile Tube Applications for Construction of the Longest Sea-Crossing Bridge in
Vietnam. Journal Terra et Aqua, IADC, The Netherlands, No. 142, March 2016, pp. 5-16.
Yee, T.W., Zengerink, E., and Choi, J.C. (2007). Geotextile tube application for Incheon Bridge Project.
Conference on CEDA Dredging Days 2007, Rotterdam, The Netherlands, CD Vol.
Yee, T.W., and Choi, J.C. (2008). Geotextile tube application for Ilsan Grand Bridge Project, Korea. First
Pan American Geosynthetics Conference and Exhibition, Cancun, Mexico, pp. 712-719.
Yee, T.W., Choi, J.C., and Zengerink, E. (2010). Revisiting the geotextile tubes at Incheon Bridge Project,
Korea. Proceedings of the Ninth International Conference on Geosynthetics, Guaruja, Brazil, Vol. 3,
pp. 1217-1220.