OVERVIEW and DESIGN ASPECTS OF REVETMENTS and GEOSYSTEMS

by
Krystian W. Pilarczyk1

ABSTRACT

Geotextile systems such as bags, mattresses, tubes and containers filled with sand or mortar can be a
good and often cheaper alternative for more traditional materials/systems as rock, concrete units or
asphalt. These new systems were applied successfully in a number of countries and they deserve to
be applied on a larger scale. Because of low price and easy construction these systems can be a good
alternative for coastal protection and coastal structures in developing countries. The main obstacle
in their application is however the lack of proper design criteria. An overview is presented on
applications and stability criteria of the existing geosystems and their limitations.

1. INTRODUCTION

Various structures/systems can be of use in hydraulic and coastal engineering, from traditional
rubble or concrete systems to more novel methods as geosystems and others. Within the scope of
the research on the stability of rock and block revetments, much knowledge has been developed
about the possible failure mechanisms and methodology on development of stability criteria under
current and wave load (CUR/RWS, 1995a,b). Until recently, no or unsatisfactory design tools were
available for a number of other (open) types of revetment and geosystems. This is why the design
methodology for block revetments has recently been extended in applicability by means of a desk-
study for a number of geosystems, such as sandbags and sand- and mortar-filled mattresses and
tubes/containers. Also other stability aspects, such as soil-mechanical stability and residual strength
were taken under consideration.
Geotextile systems utilize a high strength synthetic fabric as a form for casting large units by
filling them by sand or mortar, or as curtains collecting sand. At this moment there is a relative large
number of products of this type on the market provided by some specialistic companies all over the
world. Mattresses are mainly applied as slope and bed protection. Bags are also suitable for slope
protection and retaining walls or toe protection but the main application is construction of groins,
perched beaches and offshore breakwaters. The tubes and containers are mainly applicable for
construction of groins, perched beaches and offshore breakwaters. They can form an individual
structure conforming functional requirements for the project or as a component of the main
structure. In general, the sand-filled structure can be used as: temporary structures to learn the
natural interactins/responses, permanent structures at locations with relatively low wave attack (H <
1.5m), or submerged structures where direct wave forces are reduced. The mortar-filled systems can
resist much higher wave and current loading and, if necessary, can be interconnected by bars or by
creating a special interlocking shape.
The main advantages of these systems in comparison with more traditional methods are: a
reduction in work volume, a reduction in execution time, a reduction in cost, a use of local
materials, a low-skilled labour requirement and possibility of using of locally available equipment.
This paper aims at giving a summary of the increased knowledge, especially that concerning the
stability criteria for sand- and mortar-filled mattresses, bags, geotubes and geocontainers that have
been made available.

Noot 1 Manager R&D, Road and Hydraulic Engineering Division, Rykswaterstaat, P.O.Box 5044, 2600
GA Delft, The Netherlands, Fax +31-15-2518555, E-mail: k.w.pilarczyk@dww.rws.minvenw.nl

1
2. STABILITY CRITERIA FOR ALTERNATIVE REVETMENTS

2.1 Wave-load stability

There are two practical design methods available: the black-box model and the analytical model. In
both cases, the final form of the design method can be presented as a critical relation of the load
compared to strength, depending on the type of wave attack. For revetments, the basic form of this
relation is:
 Hs  F cos α  H 
  = with maximum  s  = 8.0 ; ctgα ≥ 2 (1)
 ∆ D  cr b
ξ op  ∆ D  cr
In which: b = exponent; b = 0.5 for riprap and gabions, b = 2/3 for block revetments and
geomattresses, and b = 1.0 for slabs, F = revetment (stability) factor (-); F =2.25 for riprap, F ≤ 4 for
geomattresses, F = 4.5 for block revetments, F = 6 for cabled block mats and F = 8 to 10 for slabs,
Hs = significant wave height (m), ∆ = ρs/ρw-1 = relative density (-), ρs = density of the protection
material, ρw = density of water (kg/m3), D = thickness of the top layer (m); (for rock: D = Dn =
(M50/rs)0.33 or Dn = 0.84 D50 ), and ξop = breaker parameter (-). For porous top layers, such as sand
mattresses and gabions, the relative density of the top layer must be determined, including the
water-filled pores: ∆ m = (1 - n) • ∆
(2)
In which: ∆ m 1= relative density including pores (-) and n = porosity of the top layer material (-). The
breaker parameter is defined as follows:
tan α
ξ op =
Hs / Lop
(3)
In which: α = slope angle (°), Lop = 1.56 Tp2 = deep-water wavelength at the peak period (m), and Tp
= wave period at the peak of the spectrum (s).
The advantage of this black-box design formula is its simplicity. The disadvantage, however, is that
the value of F is known only very roughly for many types of structures.

Figure 1 Pressure development in a revetment structure

The analytical model is based on the theory for placed stone revetments on a granular filter (Figure
1). In this calculation model, a large number of physical aspects are taken into account. In short, in
the analytical model nearly all physical parameters that are relevant to the stability have been
incorporated in the "leakage length" factor. The final result of the analytical model may, for that
matter, again be presented as a relation such as Eq. 1 where F = f( Λ ). For systems on a filter layer,
the leakage length is given as:

2
 bf D k f Λ k f bf
Λ = or = (4)
k′ D k′ D
with: Λ = leakage length (m), bf = thickness of the filter layer (m), kf = permeability of the filter or
subsoil (m/s), and k ′ = permeability of the top layer (m/s).

With a system without a filter layer (a system placed directly on sand or clay) the permeability
of the subsoil (eventually with gullies/surface channels) is filled in. For the thickness of the filter
layer it is examined to which depth changes at the surface affect the subsoil. One can fill in 0.5 m
for sand and 0.1 m for clay. The values for D and ∆ depend on the type of revetment. When
schematically representing a block on a geotextile on a gully on sand, the block should be regarded
as the top layer and the combination of the geotextile and the small gully as the filter layer. The
leakage length can be calculated using:
( k f d g + k g Tg ) D
Λ =
k′ (5)
with: kf = permeability of the filter layer (gully) (m/s), dg = gully depth (m), kg = permeability of the
geotextile (m/s), Tg = thickness of the geotextile (m), D = thickness of the top layer (m), and k ′ =
permeability of the top layer (m/s).
To be able to apply the design method for placed stone revetments under wave load to other
systems, the following items may be adapted:
· the revetment parameter F;
· the (representative) strength parameters ∆ and D;
· the design wave height Hs;
· the (representative) leakage length Λ;
· the increase factor Γ (friction, interlocking) on the strength.
Only suchlike adaptations are presented in this summarizing paper. The basic formulas of the
analytical model are given in (CUR/ RWS, 1995a and Pilarczyk, 1998).

2.2 Flow-load stability

Severe flow attack may in practice occur on revetments, such as with flow over a steep slope and
flow attack near many kinds of structures (downstream of sills, gates, discharge structures and the
like). At these structures, the flow is often specifically determined by the geometry and the boundary
conditions. With flow over a steep slope, such as on the downstream slope of a over-flow dam or
dike, the situation is less ambiguous.
When the flow velocity is known well, or can be calculated reasonably accurately, Pilarczyk's
relation (1990) is applicable:
Φ KT K h u 2cr
∆ D = 0.035 (6)
Ψ Ks 2 g
in which: ∆ = relative density (-), D = characteristic dimension/thickness (m); for rock D = Dn, g =
acceleration of gravity (g=9.81 m/s2), ucr = critical vertically-averaged flow velocity (m/s), Φ =
stability parameter (-), Ψ = critical Shields parameter (-), KT = turbulence factor (-), Kh = depth
parameter (-), and Ks = slope parameter (-).

* The stability parameter Φ depends on the application. Some guide values are given below.
Continuous toplayer Edges and transitions
Riprap and
placed blocks Φ =1.0 1.5
Mattresses, gabions,
and washed-in blocks 0.5 to 0.75 0.75 to 1.0

3
* With the critical Shields parameter Ψ the type of material can be taken into account. Some guide
values are given below.
Revetment type: Ψ (-)
riprap 0.035
loose, placed blocks 0.05
blockmats 0.07
gabions 0.07
sand and concrete mattresses 0.07

* The degree of turbulence can be taken into account with the turbulence factor KT. Some guide
values for KT are given below.
Situation: KT (-)
Normal turbulence: 1.0
Increased turbulence (i.e. river bends) 1.5
Heavy turbulence (i.e. hydraulic jump) 2.0
Sharp river bends; R/w<5: 2.0 to 2.5; (R= radius and w=width of a river)
Load due to water (screw) jet: 3.0 to 4.0

* With the depth parameter Kh, the water depth is taken into account, which is necessary to translate
the depth-averaged flow velocity into the flow velocity just above the revetment. The depth
parameter also depends on the measure of development of the flow profile and the roughness of the
revetment. The recommended formulas can be found in (Pilarczyk, 1990, Klein Breteler, 1996).
The following formulas are recommended:
2
- developed profile (logarithmic velocity profile): Kh = 2
  12 h  
 log 
  ks  
-0.2
 h
- nondeveloped profile: Kh =  
 ks 
The following indicative values for concrete mattresses and block revetments, for the water depth h
> 2 m are given below:
- developed profile: Kh = 0.2 (logarithmic velocity profile)
- nondeveloped profile: Kh = 0.4
For shallow water and rough flow (h/ks<5): Kh = 1.0. In which h = water depth (m) and ks =
equivalent roughness according to Nikuradse (m); ks ≈ 0.05 m for mattresses and smooth blocks.

* Slope parameter Ks. The stability of revetment elements also depends on the gradient under which
the revetment is applied, in relation to the angle of internal friction of the revetment. This effect on
the stability is taken into account with the slope parameter Ks, which is defined as follows:
2
 sin α 
Ks = 1-  
 sin θ 
(7)
with: θ = angle of internal friction of the revetment material (°) (about 90° for concrete mattresses
and 30 to 40°for sand-filled systems) and α = transversal slope of the bank (°).
With a downward flow along a steep slope it is difficult to determine or predict the flow velocity
exactly, because the flow is very irregular (high turbulence, inclusion of air as a result of which the
water level cannot be determined very well, etcetera). One is confronted with this when
dimensioning the revetment of (the crest and) the inner slope of a dike in the case of flooding. In
that case a design formula based on the discharge is preferable (Pilarczyk, 1998).

4
2.3 Soil-Mechanical Stability

The water movement on a revetment structures can also affect the subsoil, especially when this
consists of sand. This effect is treated within the framework of the soil-mechanical aspects and can
be of importance to the stability of the structure.
There are three aspects that will be discussed within the framework of soil-mechanical aspects:
· elastic storage;
· softening (liquefaction);
· drop in the water level.

These aspects and the accompanying damage mechanisms en design methods are discussed in detail
below. Background information can be found CUR/RWS (1995a).
Elastic storage in the subsoil is connected with the permeability and stiffness of the grain skeleton
and the compressibility of the pore water (the mixture of water and air in the pores of the grain
skeleton). Because of these characteristics, wave pressures on the top layer are passed on delayed
and damped to the subsoil of the revetment construction and to deeper layers (as seen perpendicular
to the slope) of the subsoil.
This phenomenon takes place over a larger distance or depth as the grain skeleton and the pore
water are stiffer. If the subsoil is soft or the pore water more compressible (because of the presence
of small air bubbles) the compressibility of the system increases and large damping of the water
pressures over a short distance may occur. Because of this, alternately water undertension and
overtension may develop in the subsoil and corresponding to this an increasing and decreasing grain
pressure.
Elastic storage can lead to the following damage mechanisms:
· lifting of the top layer;
· sliding of the top layer;
· sliding of the subsoil.

For the stability of the top layer, elastic storage could particularly be of importance if the top layer is
placed directly on the subsoil without there being small gullies under the top layer and, if the
permeability of the top layer is (locally) less than that of the subsoil. These conditions imply that the
leakage-length approach according to the analytical model for the stability under wave load cannot
be applied.
The stability of the subsoil may be jeopardized if,
because of elastic storage, the grain tension
decreases so strongly that insufficient sheer stress
can be absorbed in the subsoil to prevent sliding
(Figure 2). The design method with regard to the
different damage mechanisms connected with
elastic storage are presented in the form of design
diagrams. In these diagrams the permissible wave
height is plotted against the thickness of the top
layer and the slope gradient. Figure 2 Possible local sliding in the base

If the revetment consists of a top layer on a filter layer, the thickness of the filter layer should in
these diagrams be added to the thickness of the top layer (= equivalent thickness, see example in
Figure 3). Also through cyclic generation of water tension, water overtensions may occur in the
subsoil, but with impermeable top layers also directly under the top layer. In sand, these
watertensions can be calculated using the MCYCLE program developed by Delft Geotechnics. As
the top layer becomes more impermeable, the water tension manifests itself closer to the surface of
the slope. In the case of a very permeable top layer this is exactly the opposite. Softening

5
(liquefaction) can be defined as follows:
A cyclic variable load causes compaction to occur in a layer of sand. This leads to a decrease in the
pore space. The water in the pores is subjected to pressure and will want to run off. At first, water
overtension occurs. This causes a decrease in the contact pressure between the grains and with this
the resistance to sliding. Finally, the water overtension might become so large that the contact
pressure between the grains falls away completely. This is called softening or liquefaction.
The difference between liquefaction and elastic storage is that with liquefaction, water
overtension is connected with a plastic deformation of a grain skeleton instead of an elastic
deformation. Water overtension through softening occurs when the grain skeleton deforms
plastically to a denser packing. From which follows that the dangers connected with liquefaction are
smaller as the subsoil is compacted better during construction.
With regard to liquefaction, the following design rules are suggested for constructions with a
reasonably compacted subsoil:
· With a top layer on sand there is no danger of liquefaction, if:
— the slope gradient is gentler than or equal to 1:3,
— the slope gradient is gentler than 1:2 and the wave height Hs is smaller than 2 m, or
— the slope gradient is gentler than 1:2 and the subsoil is well-compacted.
· With a top layer on clay there is no danger of liquefaction.
· With a top layer on a granular filter there is generally no danger of liquefaction.

In these design rules hardly any distinction is made between types of revetment.
Through a drop in the water level a difference in the rise over the top layer may occur. A drop in the
water level may occur as a result of tide or a ship passing through a waterway or canal. As with
packed stone revetments, this is only a problem if any possible filter layer and the top layer are
sanded up and because of this obtain a low permeability.
No calculations need to be made on this phenomenon if applies:
Λsinα/2 ≤ ∆D cosα (8)

in which: Λ = leakage length (m), α = slope angle (°), ∆ = (representative) relative density of the top
layer (-), D = (representative) thickness of the top layer (m).

Figure 3 Example of geotechnical stability criteria for geomattresses /block mats; Deq=D+b/∆

6
3. CONCRETE MATTRESSES

Characteristic of concrete mattresses are the two geotextiles with concrete or cement between them.
The geotextiles can be connected to each other in many patterns, which results in each mattress
system having its own appearance and properties. An example is given in Figure 4.
The permeability of the mattress is one of the factors that determine the stability.
It is found that the permeability given by the suppliers
is often the permeability of the geotextile, or of the so-
called Filter Points. In both cases, the permeability of
the whole mattress is much smaller. A high
permeability of the mattress ensures that any possible
pressure build-up under the mattress can flow away, as
a result of which the differential pressures across the
mattress remain smaller. The stability is therefore the
largest with a large mattress permeability. In the long
term, however, pollution of the Filter Points or the
clogging of the geotextile can cause a decrease in the
permeability (Pilarczyk, 1998).

Figure 4 Example of concrete mattress

3.1 Design rules with regard to wave load

In the design rules for concrete mattresses with regard

to wave load the calculation of the leakage length is
adapted. This consideration, which is
closely related to a consideration in accordance with
the analytical model, results in a design formula in the
form of the black-box formula.
During wave attack, the mattress will be exposed to a
differential pressure which is directed upwards, as also
is the case with packed stone revetments. This takes
place the moment the wave has drawn back, just before
the wave impact. Just as with packed stone revetments, the leakage length for this differential
pressure is the most important construction-descriptive parameter. The leakage length (Λ) can be
calculated using Eqs. 4 or 5. The values of the leakage length may vary from about 0.5 to 10 m
depending on the type (permeability) of the mattress, the permeability and thickness of the filter,
and the presence of cavities under the mattress. The failure mechanism of the concrete mattress is
probably as follows:
· First, cavities under the mattress will form as a result of uneven subsidence of the subsoil.
The mattress is rigid and spans the cavities.
· With large spans, wave impacts may cause the concrete to crack and the spans to collapse.
This results in a mattress consisting of concrete slabs which are coupled by means of the
geotextile.
· With sufficiently high waves, an upward pressure difference over the mattress will occur
during wave run-down, which lifts the mattress.
· The pumping action of these movements will cause the subsoil to migrate, as a result of
which an S-profile will form and the revetment will collapse completely.

The value of stability factor F in the design formula of the black-box model (Eq. 1) depends on the
leakage length and the subsoil: F = 2 to 4. A permeable mattress on sand has a medium-sized or
small leakage length and then the value of F is 3 to 4 (max. 5). A low-permeable mattress on a filter

7
has a large leakage length and therefore an F-value of 2 to 3 (Figure 5). For the determination of the
leakage length, one is referred to the analytical model. The representative relative density ∆ follows
from the standard definition. For the representative thickness D, the average thickness should be
filled in.

It can be concluded that, compared to the available

data in literature, the derived stability relations
give a safe estimation of the stability. Because the
relations have not been verified sufficiently yet, it
is not recommended to decrease the existing
safety. In the long run, the permeability of the top
layer may diminish as a result of accretion and
silting-up. This will have a negative effect on the
stability, especially with systems with a leakage
length smaller than approximately 2 m. If the
leakage length is larger than 2 m, the effect of the
permeability of the top layer on the stability is
rather small.

Figure 5 Calculation results for concrete mattresses

3.2 Design rules with regard to flow load

A number of characteristic values for the critical flow velocity for concrete mattresses is given
below.
Thickness on slope on bottom
50 mm 2.7 m/s 3.3 m/s
100 mm 3.9 m/s 4.7 m/s
200 mm 5.5 m/s 6.4 m/s

For the application of the design formula (Eq. 6), guide values for the constants are given in Section
2.2. For the representative thickness D, one should fill in the average thickness of the top layer.

3.3 Design rules with regard to soil-mechanical stability

The flow through a concrete mattress is concentrated in the Filter Points. The permeability of the
systems filled with concrete lies approximately between 1⋅10-4 and 5⋅10-3 m/s. A concrete mattress
is less flexible than a sand mattress and does not connect to the subsoil as well as a sand mattress. In
contrast with sand mattresses, it is assumed that only the sliding of the whole mattress can occur and
not just part of it.

* Elastic storage
With regard to elastic storage, the following design example for a wave height H = 1 m and a slope
1 on 3 is given. The required thickness of the mattress on sand for various failure mechanisms and
wave steepnees (Sop) is equal to:
Failure type Sop = 0.03 Sop = 0.05
Lifting of toplayer 0.35 m 0.25 m
Partial sliding toplayer 0.80 m 0.60 m
Sliding of toplayer 0.30 m 0.25 m
Sliding of subsoil 0.55 m 0.40 m

Concrete mattresses are mostly stiff and anchored at the top. Therefore, not the sliding and/or

8
uplifting of the toplayer but the sliding of the subsoil is the most dangerous (for H=1m and
Sop=0.05, the required total thickness (top layer + filter) is 0.40m).

If the systems are placed on a filter, one can take into account an increase in the stability with regard
to elastic storage. For the total thickness of a concrete mattress on a filter, D + b/∆t can be filled in,
where b is the thickness of the filter (= equivalent thickness, Figure 3).

* Liquefaction
The design rules with regard to liquefaction do not differ from those presented in Section 2.3.

4. SAND MATTRESSES

A sand mattress consists of two geotextiles attached onto each other, between which sand is
interposed. This way, a mattress is formed of sausages lying next to each other which run from the
top to the bottom of the slope and which are interconnected. The lower geotextile is usually flat and
the upper geotextile lies on top of it, in arches.

* Construction/design/repair
The edges and connections of sand mattresses are vulnerable and must therefore be finished
carefully. Mattresses lying next to each other can be sown together and the ends can be secured with
for example ground anchors. In actual practice, mattresses are not only threatened by the hydraulic
load. The possibility of vandalism occurring, limits sand mattresses to being applied in places where
unauthorized persons do not have access to. The system is also vulnerable to collision, (drifting) ice,
floating bulky refuse, sunlight and chemical degradation.

* Stability
Sand mattresses cannot be used when the significant wave height Hs is larger than 1.0 m (max. 1.5
m in case of properly compacted subsoil). Unfortunately, not much research has been conducted
into the stability of sand mattresses. Besides Pilarczyk's design formula (1990), a small-scale model
investigation, a desk study and a prototype experiment have been found. Based on these, the
following value for A in the design formula (Eq. 1) is recommended according to the black-box
model: F = 4 to 5. In this formula the relative density including pores ∆m should be filled in for the
representative relative density. For the representative thickness D of the mattress, the average
thickness should be filled in:

D
= 0.7 to 0.8
100% filled : Dd

D
90% filled : = 0.6 to 0.7
Dd

with: D = average thickness of the mattress (m) and Dd = max. diameter of the sausages (m).

Above a flow velocity of 1.5 m/s (max. 2 m/s), the sand in the mattresses is no longer internally
stable. The design formula on stability against currents is given in section 2.2.
A sand mattress is relatively flexible and connects closely to the subsoil. The geotechnical design
criteria are similar to those for the concrete mattresses.

9
5. GEOBAGS, GEOTUBES AND GEOCONTAINERS

Conventional breakwaters with rock and concrete units have a long history and much experience has
been gained on their design and construction (CUR/RWS, 1995). However, the increasing need for
more economical designs and shortage of natural rock in certain geographical areas have stimulated
in recent years the alternative designs utilizing geosystems and local materials.
Large geobags, geotubes and geocontainers hydraulically and/or mechanically filled with
(dredged) materials have been successfully applied in hydraulic and coastal engineering in recent
years (shore protection, breakwaters, etc.) (Pilarczyk, 1995, 1999, Pilarczyk & Zeidler, 1996,
Fowler et al, 1995). Especially for creating the perched beaches the sand bags and/or sand tubes can
be an ideal, low-cost solution for constructing the submerged sill (with a low or moderate wave
loading). Some concepts are shown in Figure 6 (Tetra Tech, 1982).
They can also be used to store and isolate contaminated materials from harbour dredging, and/or to
use these units as bunds for reclamation works. Also, the geocurtains can be applied for
construction of submerged sills and reefs. A brief review of these systems can be found in
(Pilarczyk, 1995b).
Until very recently, geosystems were mostly applied as temporary structures. The reason for that
was their relatively low resistance to the hydraulic loadings (waves and currents), the lack of proper
design criteria, and low durability in respect to UV-radiation and vandalism. For exposed
applications, the aesthetic aspects of geosynthetics may also play a role.

Figure 6 Some concepts of application of geosystems

In general, it can be said that all coastal protection systems and materials have some advantages
and disadvantages which should be recognized before the choice is made. There is no one ideal
protective system or material. Each material and system has a certain application at certain loading
conditions and specific functional requirements for the specific problem and/or structural solution
(Pilarczyk & Zeidler, 1996).
In the past, the design of geotextile systems for various coastal applications was based mostly on
rather vague experience than on the general valid calculation methods. However, the increased
demand in recent years for reliable design methods for protective structures have led to the

10
application of new materials and systems (incl. geotextile systems) and to research concerning the
design of these systems. Contrary to research on rock and concrete units, there has been no
systematic research on the design and stability of geotextile systems. However, past and recent
research in The Netherlands, USA and in some other countries on a number of selected geotextile
products has provided some results which can be of use in preparing a set of preliminary design
guidelines for the geotextile systems under current and wave attack (Den Adel, 1996, Klein
Breteler, 1996, Pilarczyk, 1995, 1996, 1997, 1998, 1999, Wouters, 1995, Bezuijen, 1998).
In the next sections the large fill-containing geosystems (geotubes and geocontainers filled with
sand or mortar) and their design aspects (including structural and hydraulic stability) are discussed.

5.1 Description of geosystems applicable for breakwaters

Geobags
Geobags can be filled with sand or gravel (or cement, perhaps). The bags may have different shapes
and sizes, varying from the well-known sandbags for emergency dikes to large flat shapes or
elongated "sausages" (see Figure 7). The most common use for sandbags in hydraulic engineering is
for temporary structures (Wouters, 1995).
Uses for sand- or cement-filled bags are, among
other things:
· repair works (see Figure 7);
· revetments of relatively gentle slopes and
toe constructions;
· temporary or permanent groynes and
offshore breakwaters;
· temporary dikes surrounding dredged
material containment areas.

Figure 7 Application of geobags

Because this material is easy to use and cheap, it

is extremely suitable for temporary structures. A
training groyne is a good example. The working
of a groyne is difficult to predict in advance.
That is why it is a good procedure to make such
a construction using a relatively cheap product first, to see how one thing and another works out,
and subsequently either make improvements or, after some time, a permanent structure. Above a
flow velocity of 1.5 m/s, the geosystems filled with sand cannot be used because the sand in the
systems is no longer internally stable.
Sandbags can be placed as follows:
1. As a blanket: One or two layers of bags placed directly on the slope. An "interlocking" problem
arises if the bags are filled completely. The bags are then too round. A solution is not to fill the
bags completely, so that the sides flatten out somewhat, as a result of which the contact area
becomes larger.
2. As a stack: Bags stacked up in the shape of a pyramid. The bags lie halfoverlapping with the
long side parallel to the shoreline.

When installing geosystems, one should see to it that this does not take place on a rough foundation.
Sharp elements may easily damage the casing of the element. Geosystems must not be filled
completely. With a fill ratio of approximately 75% an optimum stability of the elements is reached.
A sound soil protection is necessary if gravel (sand) sausages are used in circumstances where they
are under attack of flow or waves. Stability criteria are given in Pilarczyk (1999).

11
Tube system
Geotube is a sand/dredged material filled geotextile tube made of permeable but soil-tight
geotextile. The desired diameter and length are project specific and only limited by installation
possibilities and site conditions. The tube is delivered to the site rolled up on a steel pipe. Inlets and
outlets are regularly spaced along the length of the tube. The tube is filled with dredged material
pumped as a water-soil mixture (commonly a slurry of 1 on 4) using a suction dredge delivery line
(Figure 8). The choice of geotextile depends on characteristics of fill material. The tube will achieve
its desired shape when filled up to about 80%; a higher filling grade is possible but it diminish the
friction resistance between the tubes.

Figure 8 Filling procedure of Geotube

Tube can be filled on land (e.g. as dikes for land reclamation, bunds, toe protection or groyns) or in
water (e.g. offshore breakwaters, sills of perched beaches, dikes for artificial islands or interruption
of gullies caused by (tidal)currents). The tube is rolled out along the intended alignment with
inlets/outlets centered on top. When a tube is to be placed in water, the effects of buoyancy on the
tube geotextile prior to filling as well as on the dredged material's settling characteristics must be
considered. In order to maximize inlet/outlet spacing, an outlet distant from the inlet may be used to
enhance the discharge of dredged slurry and thereby encourage and regulate the flow of fill material
through the tube so that sufficient fill will flow to distant points. The tube will achieve its desired
shape when it is filled up to about 70 to 80% of the theoretical circular diameter or a height equal to
about a half of the flat width of the tube; a higher filling grade is possible but it diminishes the
frictional resistance between the tubes.

Commonly, the filter geotextile (against scour) and flat tube are fully deployed by floating and
holding them in position prior to beginning the filling operation. The filter geotextile is often
furnished with small tubes at the edges when filled with sand holds the filter apron at place. This
apron must also extend in front and behind the unit, commonly 2 times the height of the entire
structure or 2 to 3 times the local wave height.

Container system
Geocontainer is a mechanically-filled geotextile and "box" or" pillow" shaped unit made of a soil
tight geotextile. The containers are partially prefabricated by sawing mill widths of the appropriate
length toegether and at at the ends to form an elongated "box". The "box" is then closed in the field,
after filling, using a sewing machine and specially designed seams (Figure 9). Barge placement of
the site-fabricated containers is accomplish using a specially configured barge-mounted crane or by
bottom dump hoppers scows, or split barges. The containers are filled and fabricated on the barge
and placed when securely moored in the desired position. Positioning of barge for consistent

12
placement - a critical element of constructing "stacked" underwater structures - is accomplished
with the assistance of modern surveying technology. The volume of applied geocontainers was up to
1000 m3.

Figure 9 Filling and placing of geocontainer

The advantage of these large barge-placed geocontainers include:

* Containers can be filled with locally available soil which may be available from simultaneous
dredging activities;
* Containers can be relatively accurately placed regardless of weather conditions, current

13
 velocities, tides, or water depths;
* Contained material is not subject to erosion after placing;
* Containers can provide a relatively quick system build-up;
* Containers are very cost competitive (for larger works).

5.2 Design considerations

When applying geobags, geotubes and geocontainers the major design considerations/problems are
related to the integrity of the units during filling, release and placement impact (impact resistance,
seam strength, burst, abrasion, durability), the accuracy of placement on the bottom (especially at
large depths), and the stability under current and wave attack.

The geotextile fabric used to construct the tubes is designed to:

* contain sufficient permeability to relieve excess water pressure,
* retained the fill-material,
* resist the pressures of filling and the active loads without seams or fabric rapture,
* resist erosive forces during filling operations,
* resist puncture, tearing, and ultraviolet light.

The following design aspects are particularly of importance for the design of containers:
a) Change of shape of units in function of perimeter of unit, fill-grade, and opening of split barge,
b) Fall-velocity/equilibrium velocity, velocity at bottom impact,
c) Description of dumping process and impact forces,
d) Stresses in geotextile during impact and reshaping,
e) Resulting structural and executional requirements, and
f) Hydraulic stability of structure.

Some of these aspects are briefly discussed hereafter.

5.3 Shape and mechanical strength of geotubes

For the selection of the strength of the geotextile and calculation of a required number of tubes for a
given height of structure, knowledge of the real shape of the tube after filling and placing is
necessary. The change of the cross-section of the tube depends on the static head of the (sand)slurry.
Depending on this static head, the laying method and the behaviour of the fill-material inside the
tube, it is possible that the cross-sectional shape of the filled tube (with a theoretical cross-section
with a certain diameter, D) will vary from a very flat hump to a nearly fully circular cross-section.
More recently, Liu (1978, 1981), Silvester (1985, 1990), Carroll (1994), Kazimierowicz (1994), and
Leshchinsky (1995, 1996) prepared some analytical or numerical solutions and graphs allowing the
determination of the shape of sand- or mortar-filled tubes based on some experiments with water.
The Leshchinsky's method (PC-program) combines all the previous developments and can be
treated as a design tool.

The design of the shape of the geotube is an iterative process. To obtain a proper stability of the
geotube and to fulfill the functional requirements (i.e. required reduction of incoming waves/proper
transmission coefficient the width and the height of the tube (= a certain crest level) must be
calculated. If the obtained shape of geotube does not fulfill these requirements a new (larger) size of
a geotube must be taken into account or a double-line of tubes can be used. As an example the
shapes of the geotube (with the theoretical diameter of 3.25m) for the height of 1.8 m, 2.0 m and 2.1
m, based on the Leshchinsky's calculation model, is shown below (see also Figure 10). The maxim-
um width is b = 4.15 m (d=1.8 and 2.0m) to 4.0 m (d=2.1m), and the cross-section area is A = 6.41,
6.88, and 7.06 m2 respectively. The width of base of the tube (= contact width with foundation

14
layer) is about 3.10, 2.90, and 2.60 m respectively.
The required minimum pumping pressure
is about 0.2 psi = 1.5 kPa = 0.15 m of a
water column for d = 1.8 to 2.0m (may be it
is too low for practical realization?). In
case of d = 2.1 m the minimum pumping
pressure is 0.4 psi = 3.0 kPa = 0.30 of a
water column. The required tensile strength
of geotextile is about 80 kN/m (including
safety factors).

Figure 10 Shapes of geotube with 3.25m dia.

5.4 Dumping process of containers and practical uncertainties

A summary of various forces during the dumping and placement process is given in Figure 11
(Bezuijen, 1998, Pilarczyk, 1999).

1. The required perimeter of geotextile sheet must be sufficient enough to release geocontainer
through the given split width bo for a required cross-sectional area of material in the bin of barge Af
(or filling-ratio of fill-material in respect to the max. theoretical cross-section). The derivation of the
required minimum length of perimeter of geotextile sheet is given in (Bezuijen, 1998 and Pilarczyk,
1999). After opening of the split of a barge the geo container is pulled out by the weight of soil but
at the same time the friction forces along the bin side are retarding this process. Due to these forces
the tension in geotextile is developing at lower part and both sides of the geocontainer. The upper
part is free of tension till the moment of complete releasing of geocontainer. The question is how far
we are able to model a friction and the release process of geocontainer.

Figure 11 Development of forces during dumping of geocontainers

15
Friction and tensile forces in the geotextile during the release of the geocontainer

Figure 12 The schematization of a bin for the calculation of forces

Capacity (volume) of the geocontainer:

V = 0.5 L b h
where: L = total length of the geocontainer, b =
top width , and h = depth of soil.
The touch (side) length at the bottom 'l' is equal
to: l = _ (h2 + (0.5 b)2)
and the split-opening width bo as a function of the
radius of rotation (R) and the angle of opening
(Θ) is given by: bo = 2 R sinΘ.

Wa = bo h γs

Wb = 0.25 (b - bo) h γs = 0.25 h2 ctgΘo γs

h = 0.5 (b - bo) tanΘo

Af= V/L= 0.5(b + bo) h = 0.25(b2 - bo2) tanΘo

or
b = _(4 Af ctgΘo + bo2)

The equilibrium of the rectangular part in the

vertical direction is:

Tv + Potanφ = 0.5(Wa- boww)

or
Tv = 0.5(Wa- boww) - Potanφ
and
Figure 13 Acting forces in the bin
T = {0.5(Wa- boww) - Potanφ} cosecΘo (9)

16
The equilibrium of the triangle in the direction of the friction force F is:

F= WbsinΘo - PocosΘo + Po tanφ sinΘo + T (10a)

The equilibrium of the triangle in the direction of the normal force N is:

N = Wb cosΘo + Po sinΘo + Po tanφ cosΘo (10b)

where Po is the static earth pressure equal to: Po = 0.5 K γs h2 and ww is the water pressure in the
function of the loaded draft db, F is the friction force, T is the tensile force in the geotextile, K = the
coefficient of static earth pressure, and φ the angle of internal friction. The criterion is that the
friction force F cannot exceed µN in which µ is the friction coefficient between the geocontainer
and the bin of the barge. Thus, Fmax = µ N. The moment that F exceeds Fmax, sliding along the barge
starts and the maximum tensile force is reached.

2. Geocontainer will always contain a certain amount of air in the pores of soil and between the soil
and the top of (surplus) geotextile providing an additional buoyancy during sinking. The amount
and location of air pockets depends on soil consistency (dry, saturated) and uniformity of dumping.
The air pockets will exert certain forces on geotextile and will influence the way of sinking. The
question is how to model in a proper way the influence of soil consistency and air content on shape
and stresses in geotubes/geocontainers.

3. The forces due to the impact with the bottom will be influenced by a number of factors:
* consistency of soil inside the geocontainer (dry, semi-dry, saturated, cohesie, etc.) and its
physical characteristics (i.e. internal friction);
* amount of air;
* permeability/airtightness of geotextile;
* strength characteristics of geotextile (elasticity/elongation vs. stresses, etc.);
* fall-velocity (influenced by consistency of soil; saturated soil diminish amount of air but increa-
ses fall speed);
* shape and catching surface of geocontainer at impact incl. effect of not horizontal sinking (i.e.
catching of bottom with one end);
* type of bottom (sand, clay, soft soil, rock, soil covered with rockfill mattress, etc.) and/or type
of sublayer (i.e. layer of previous placed containers).

During the impact the cross-sectional shape of geocontainer will be undergoing a continous
reshaping; from cone shape, first probably into a transitional cylindrical shape, and through a certain
relaxation, into a semi-oval shape or flat triangular/rectangular shape dictated by soil type, perimeter
and elongation characteristics of geotextile. The question is how far we are able to model this
impact phenomena and resulting forces/stresses in geotextile. The impact forces with the bottom are
a function of the fall velocity (dump velocity) of a geocontainer. The derivation of the fall velocity
is described hereafter.

* The required perimeter of the geotextile enveloping the geocontainer, must be sufficiently long to
resist the forces during the release of the geocontainer through a given split width bo, at a fixed
cross-sectional area of material in the geocontainer: Af, Figures 14 and 15.
The required perimeter of geotextile sheet, So, is given by:

 Af 
S 0 ≥ (2.5 to 4)  + b0
 b0 
(11)
The real shape on the bottom (Fig. 15) will be more close to rectangular one for low filling grade

17
(φ), while more close to the semi-oval shape for a high filling grade. Therefore the maximum height
of geocontainer, ab, in the final position will be:

S0 2 S0
4
( )
1 - 1 - 16ϕφ < a b <
π
(
1 - 1 - 14ϕφ )
(12)
φ is the filling-grade ratio (≤1) and _=Ao/So2, where Ao is the maximum area of the cross-section of
the split barge. These equations provide an estimation for the average number of geocontainers,
required for a given structure.

* When the geocontainer hits the bottom it

reshapes. The originally vertically orientated
cone transforms into a cylinder and finally
reshapes into a horizontally orientated ellipse
(Fig. 15).
• The impact forces when the
geocontainer hits the bottom are a
function of the kinetic energy and thus
the dump velocity of a geocontainer.
Next to the accelerating gravity, flow
resistance decelerates the geocontainer.
When the gravitational force equals the
flow resistance force, an equilibrium
velocity is reached. This velocity
(roughly 5 m/s), is:
•
2 Vol ( ρ s - ρ w ) g Figure 14 Figure 15
V max =
A ρ w Cd
(13)
with: Vmax = equilibrium velocity [m/s], ρs = density of the
material in the geocontainer [kg/m ], ρw = density of water [kg/m3], Vol = volume of geocontainer
3

[m3], A = flow catching surface area of the geocontainer [m2], Cd=drag coefficient [-].
* Just before touch down the geocontainer has maximum kinetic energy. This energy will be
completely dissipated when the geocontainer is at rest. The main sources for energy dissipation are
the work done by straining the geotextile, by reshaping the subsoil and by reshaping the material in
the geocontainer.

Figure 16 Schematization of impact forces

Assuming that a fraction K of the kinetic energy is dissipated by straining the geotextile, the
pressure inside the geocontainer due to the impact, qo, results in a tensile force, F, of the geotextile
(Figure 16):

18
 

F = q0 R L 



K Vol ρ s v 2 E ′ 
q0 =
S L R2 
(14)

with q0 the overpressure inside the geocontainer [N/m2], Vol the volume of the material in the
geocontainer [m3], ρs the density of the fill material [kg/m3] (ρs = 1600 for dry- and 2000 for
saturated sand), v the velocity at touch down [m/s], E' = stiffness modulus of the geotextile, +E⋅tg
[N/m], E = Young modulus of geotextile [N/m2], tg = thickness of geotextile [m], L = length of
geocontainer [m], R = radius of geocontainer (=S/2π)[m], S = perimeter of geocontainer [m], and K
= dissipation factor [-].
Other modes of failure due to the impact and reshaping and possible calculation methods can be
found in den Adel (1996) and Bezuijen (1999).

4. In final situation the geocontainers will perform as a core material of various protective structures
or as independent structure exposed to loading by currents and waves, and other loadings (ice,
debris, ship collision, vandalism, etc.). In most cases the geocontainers will be filled by fine (loosely
packed) soils. The question is how these structures will behave in practice under various types of
external and internal loadings.

Practical note: The prototype experience indicate that geocontainers with volume up to 200 m3 and
dumped in water depth exceeding 10 m have been frequently damaged (collapse of seams) using
geotextile with tensile strength lower than 75 kN/m, while nearly no damage was observed when
using the geotextile with tensile strength equal or more than 150 kN/m. This information can be of
use for the first selection of geocontainers for a specific project.

5.5 Hydraulic stability criteria for geosystems

The general calculation scheme for stability

of geotubes is shown in Figure 17 where H
= wave height and k = reflection
coefficient (Delft Hydraulics Lab., 1973).
Based on this schematisation the resistance
to the horizontal displacement and to the
overturning can be calculated. During the
hydraulic testing the both components of
stability are implicitely included in the test
results. The design method with regard to
wave load based on test results can be
presented as a critical relation of the load
compared to strength, depending on the
type of wave attack. Figure 17 Calculation scheme for geotubes

The stability relation of sand, gravel or cement bags which are used as protection elements on a
slope appears to deviate somewhat from the formula according to the black-box model. For regular
waves the recommended formula is as follows:
 H  3.5
  =
 ∆ D  cr ξo
(15)

19
In which ∆ is the relative density if the pores are completely filled with water (∆m). The
representative thickness D is the average thickness of the top layer, measured perpendicularly to the
slope. If this stability relation is combined with the relation found between Hs and H, (significant
wave height with irregular waves and the wave height with regular waves) this results in the
following stability relation:
 Hs  2.5
  =
 ∆ D  cr ξ op
(16)
For concrete sausages (tubes) used as a protection element on the crest of a low or underwater
breakwater, it is found that the following stability relation for regular waves can be used:
1/3
 H  H
  = 3.2  
 ∆ b  cr  L0 
(17)
In which b is the width of the sausage. Should two sausages be connected, the widths of both
sausages together can be filled in for b. If the sausage is placed with its longitudinal direction
perpendicularly to the axis of the breakwater, the following stability relation applies:
 H
  = 1.0
 ∆ l  cr
(18)
In which l is the length of the sausage (tube).

Concerning the flow load, above a flow velocity of 1.5 m/s, the sand in the systems is no longer
internally stable (as is more or less the case for all geosystems filled with fine material) and may
deform. The design formulas on current attack can be found in (Pilarczyk, 1998). Also for these
systems the soil-mechanical stability should be treated according to the criteria mentioned in
(Stoutjesdijk, 1996). More information on the use of geotubes in construction of breakwaters and
slope protection can be found in Delft Hydraulics (1973, 1975, 1983, 1994) and Pilarczyk (1999).
Tanaka et al (1990) and Sawaragi (1995) provide information on wave control by flexible mound.

5.6 Other design considerations (durability/UV-protection)

Durability of geotextiles is frequently coming question especially concerning the applications where
a long life-span is required. Geotextile is a relatively new product. The first applications are from
60's. Recently, we have tested in the Netherlands some 30-years old geotextiles used as a filter in
revetment structures. In general, these geotextiles were still in a good conditions (still fulfilling
properly the prescribed filter and strength functions). The technology of geotextiles is improved to a
such extend that the durability tests under laboratory condition indicate the life-time of geotextiles
at least of 100 years (when not exposed to UV radiation).
There is no problem with durability of the geosystems when they are submerged or covered by
armour layers. However, in case of exposed geosystems the UV radiation and vandalism are the
factors which must be considered during the design. All synthetics are vulnerable to UV. The speed
of UV degradation, resulting in the loss of strength, depends on the polymer used and type of
additives. Polyesters (PET) are by nature more light stable than, for example, polyamide (PA) and
polypropylene (PP). As an example, the Dutch tests with geosynthetic ropes (stabilised and not
stabilised) exposed to various environment have provided the following results (see Table)
concerning the strength of the surface yarns after 3 years (in %) in comparison with the original
strength.
To avoid the problem with light degradation the fabrics must be properly selected (i.e. polyester)
and UV stabilized (Santvoort, 1994). As the period in which the fabric is exposed is short (in terms
of months), no serious problems are to be expected. In case of more or less permanent applications
under exposed conditions the fabric must be protected against direct sunlight. There is a number of
methods of surface protection for geosystems. To provide additional UV and abrasion protection to

20
the exposed sections of tubes, a coating of elastomeric polyurethane is often used. This coating,
however, has a tendency to peel after about a number of months and therefore, has to be reapplied.

fabric type land-climate sea-climate intertidal zone 50 m under

ebb-flood the sea
PET: stabilised 63% 62% 94% 93%
PA : stabilised 33% 8% 85% 91%
PA : no 14% 6% 80% 71%
PP : stabilised 41% 46% 93% 95%
PP : no 1% 16% 92% 95%
Note: the geosynthetics under water and in the intertidal zone show very little degradation in strength in
comparison with geosynthetics placed on land; in the inter-tidal zone the geosynthetics are covered very
soon by algae which provide very good UV protection.

The permanent surface protection by riprap or blockmats (Figure 18) is a rather expensive solution
and it will normally be applied only when it is dictated by necessity due to a high wave loading or
danger of vandalism or other mechanical damage ie. boating, anchoring, etc. In other cases it will be
probably a cheaper solution to apply a temporary protection of geotextile tubes by an additional
layer of a strong geotextile provided with special UV-protection layer (Figure 19).

Figure 18 Geotube with a blockmat as armouring

Figure 19 Geotube with a surface protection layer

This geotextile layer might provide a protection for at least 10 years. Every 10-years (probably
more) a new geotextile surface-layer must be added, however, it can be that the life-time of this
layer is much longer. There is always a possibility to pass on to a permanent protection if necessary.

21
In case of this solution a maintenance program is necessary to quarantee the maintenance budget at
a proper time. To avoid lifting up, this protective layer must be prepared by using a strong, heat-
stabilized geotextile (i.e. polyester, 100kN/m), but relatively open (i.e. O90=1mm). Since a part of
the tubes is exposed permanently, this geotextile layer must be provided with an extra protection
against U.V.-radiation. This extra protection can be realized by adding the highly U.V.-stabilized
nonwoven fleece needled onto the main fabric. The function of this felt layer is also to trap the
sediment particles and algae, which give again extra U.V.-protection. The experience with this
system in Nigeria and in the Netherlands, under exposed conditions, was very satisfactorily
(Tutuarima & van Wijk, 1984). This protection layer must be properly anchored by edge (concrete-)
blocks/beams (about 0.5 to 1 tonne per meter of the tube). The edges of the geotextile should be
wrapped up (folded over) or connected to the anchor blocks (i.e. by pins). It is also possible to
anchor the edges by a heavy rock (stone class 300-1000kg, D50 = 0.75m) over the width of the edge
of about 2m (= 3 stones). There should be an additional width of the geotextile allowing the
wrapping up over and under the stones.

5.7 Geocurtains (BEROSIN CURTAINS)

The BEROSIN curtain is a flexible offshore

structure made of various woven geotextiles
which, after being placed near the shore and
anchored to the bed, catches the sand transported
by currents and waves, thus providing accretion
on a shore and preventing erosion (Figure 20). The
proper choice of permeability of a geotextile
creates the proper conditions for the sedimentation
of suspended sediment in front of or under the
curtain and at the same time allows the water to
flow out without creating high forces on the curtain and the anchors.

Figure 20 Principles of BEROSIN curtain

The quick sedimentation process will help to minimize the forces on the lower edges of the
curtain. Special open pockets on the
surface at the lower edge of a curtain
are filled with sand already at the
beginning of the sedimentation
process and function as anchors. To
allow the process of catching
sediment, the upper edges are
equipped with a floating capacity
(floaters with depth compensation)
adapted to the specific flow
conditions and the depth (Van der
Hidde, 1995, Van der Wal, 1997,
Pilarczyk, 1999). To create the larger
submerged reefs (with a sufficient
wave reduction)the BEROSIN
curtains can be placed in interlinked
sections as shown in Figure 21.
Figure 21 Installation of 2-rows BEROSIN sections

22
The horizontal system can also be applied directly on the beach (in a swash zone); the sediment
transported during the run-up phase (or during high tide) will be accumulated under the cover of
Berosin during the run-down phase (or by tidal currents during falling tide). It will ensure a very
rapid growth of the beach. In that case, however, the system can be affected by vandalism,
especially in touristic areas or by ice movement.
In the case of the coast of Vlieland in the Netherlands, some of the horizontal curtains placed
experimentally in the intertidal zone have provided a growth of the beach/foreshore of 0.5 to 1.0 m
within a week, whereas with others it took a few weeks. It was also recognized that the sheets
(curtains) can be easily damaged in the vicinity of rock, due to abrasion (one curtain was connected
to the existing rock groyne). On the other hand, the heads of the existing groynes were badly
damaged and the beach between the groynes was eroded during storms, whereas the area protected
by the curtains remained in good condition. It seems that this system can provide a low-cost
measure for steering the morphological processes. However, more prototype experiments in various
wave climates are needed before the final conclusions on the effectiveness and durability of this
system in various design conditions can be drawn.

5.8 Applications of geosystems

Nicolon, BV (1998), has copyrighted the name for GeoTubes and GeoContainers. Geotubes are
commonly used to assist in dike and groin construction whereas Geoconcainers are either dumped
from dump trucks or bottom dump (split) barges. Geotubes have been used extensively on the
northern shores of The Netherlands and Germany for dike construction with fine-grained dredged
sands pumped to form a barrier dike for subsequent hydraulic fill behind the dike and as a core for
breakwaters or dikes (Figures 22 and 23).

Figure 22 Marken breakwater (NL)

Figure 23 Geocontainers in dike construction; river Ems, Germany

23
Geocontainers have been used for construction of underwater berms and scour protection of banks.
Dredged material filled tubes have been used as containment dikes in Brazil and France and more
recently in the Netherlands for both river training structures on the rivers Waal, and Old Meuse and
as shoreline protection at Leybucht on the North Sea in Germany (Spraque, 1994). Application of
these systems is illustrated by a number of projects executed in the Netherlands and Germany
(Figure 19). Several Geotube and Geocontainer projects have recently been designed and
constructed in the U.S., Taiwan, Malaysia, Philippines and Japan, and their performance is being
documented so that improved design and construction methods can be recommended (Fowler et al,
1995, Pilarczyk, 1995, 1999, Van Oord-ACZ, 1995, Wouters, 1995, Leshchinsky, 1995, 1996). The
theoretical models to simulate the fall velocity and the impact of geocontainer on the subsoil have
been developed and calibrated with the test results (Bezuijen, 1998). However, the theoretical model
to simulate the impact give indicative results only.

6. CONCLUSIONS

The following conclusions can be drawn based on the actual developments and experience.

* Geosystems offer the advantages of simplicity in placement and constructability, cost

effectiveness, and minimal impact on the environment.
* When applying this technology the manufacturer's specifications should be followed. The
installation needs an experienced contractor.
* When applying geotubes and geocontainers the major design considerations/problems are related
to the integrity of the units during release and impact (impact resistance, seam strength, burst,
abrasion, durability etc.), the accuracy of placement on the bottom (especially at large depths), and
the stability.
* Information presented on the stability criteria will be of help in preparing the preliminary
alternative designs with geosystems.
* The geotextile systems can be a good and mostly cheaper alternative for more traditional
materials/systems. These new systems deserve to be applied on a larger scale. However, there are
still much uncertainties in the existing design methods. Therefore, further improvement of design
methods and more practical experience under various loading conditions is still needed.
* The state-of-the-art of the actual knowledge on the geosystems in hydraulic and coastal
engineering can be found in Pilarczyk (1999).

REFERENCES
CARROLL, R.P., 1994, Submerged geotextile flexible forms using non-circular cylindrical shapes,
Geotechnical Fabric Report, IFAI, St. Paul, MN, 12(8).
CUR/RWS, 1995a, Design Manual for Pitched Slope Protection. CUR report 155, A.A. Balkema.
CUR/RWS, 1995b. Manual on the use of rock in hydraulic engineering. CUR report 169, A.A. Balkema
Publisher, Rotterdam.
BEZUIJEN, A., 1998,1999, Geosystems, Delft Geotechnics.
DELFT HYDRAULICS LABORATORY, 1973, Breakwater of concrete fill hoses, Report1085.
DELFT HYDRAULICS LABORATORY, 1975, Artificial Islands in the Beaufot-sea: M 1271 part III,
comparison of stability of shore protection with gabions and sand sausages (2-dim.); M 1271 part V,
stability of shore protection with sand sausages on circular island.
DELFT HYDRAULICS LABORATORY/DELFT SOIL MECHANICAL LAB., 1983, Stability of
ProFix sand filled mattresses under wave action, Report of model investigation, R1903.
DELFT HYDRAULICS/NICOLON, 1994, Stability of breakwaters constructed with Geotubes
or Geocontainers, 2-d. model tests, Report on the model investigation, H2029.
DEN ADEL, H., 1996, Forces due to impact and deformation of geotubes (in Dutch), Delft Geotechnics,
report CO-345040.

24
DEN ADEL, H., HENDRIKSE, C.H., and PILARCZYK, K.W., 1996, Design and application of geotubes
and geocontainers, Proceedings of the 1st European Geosynthetics Conference (EuroGeo), Maastricht,
A.A. Balkema (Publisher).
FOWLER, J., D. TOUPS, Ch. MESA and P. GILBERT, 1995, Geotextile contained contaminated
dredged material, Marina del Rey, Los Angeles and port of Oakland, California, Proc. 14th World
Dredging Congress (WODA), Amsterdam.
KAZIMIEROWICZ, K., 1994, Simple Analysis of Deformation of Sand-Sausages, 5th Inter. Conf. on
Geotextiles, Geomembranes and Related Products, Singapore.
KLEIN BRETELER, M., 1996, Alternatieve open taludbekledingen (Alternative open slope revetments, in
Dutch), Delft Hydraulics, H1930.
LESHCHINSKY, Dov, and LESHCHINSKY, Ora, 1995, Geosynthetic Confined Pressurized Slurry
(GeoCops): Supplement Notes for Version 1.0, May 1995 (Nicolon/US Corps).
LESHCHINSKY, Dov, Ora LESHCHINSKY, HOE I. LING, and PAUL A. GILBERT, 1996, Geosyn-
thetic Tubes for Confining Pressurized Slurry: Some Design Aspects, Journal of Geotechnical
Engineering, ASCE, Vol. 122, No.8, August.
LIU, S,. Gen, 1978, Sand sausages for beach defence work, Univ. of Western Australia.
LIU, S., Gen, 1981, Design criteria of sand sausages for beach defences, XIX IAHR Congress, New
Delhi, India.
NICOLON, 1998, Product information, Ten Cate Nicolon bv, P.O. Box 236, 7600 AE Almelo, The
Netherlands.
PILARCZYK, K.W. (ed.), 1990. Design of seawalls and dikes - Including overview of revetments. In:
Coastal Protection, A.A. Balkema Publisher.
PILARCZYK, K.W., 1995a, Novel systems in coastal engineering; geotextile systems and other methods,
Rijkswaterstaat, Delft.
PILARCZYK, K.W., 1995b,Geotextile systems for coastal protection - an overview, Proced. of the 4th
COPEDEC, Rio de Janerio, Brazil.
PILARCZYK, K.W., 1996, Geosystems in hydraulic and coastal engineering - An overview,
Proceedings of the 1st European Geosynthetics Conference (EuroGeo), Maastricht.
PILARCZYK, K.W., 1997, Application and design aspects of geocontainers, Geosynthetics '97, USA.
PILARCZYK, K.W., 1998, Stability criteria for geosystems - an overview -, 6th Inter.
Geosynthetics Conference, Atlanta, USA.
PILARCZYK, K.W., ed., 1998, Dikes and Revetments, A.A. Balkema Publisher, Rotterdam.
PILARCZYK, K.W., 1999, Geosynthetics and geosystems in hydraulic and coastal engineering, A.A.
Balkema Publisher, Rotterdam, The Netherlands, ( www.balkema.nl).
PILARCZYK, K.W. and ZEIDLER, R.B., 1996, Offshore breakwaters and shore evolution control,
A.A. Balkema Publisher, Rotterdam.
SANTVOORT, G., 1994, Geotextiles and Geomembranes in Civil Engineering, A.A. Balkema
SAWARAGI, T., 1995. Coastal Engineering - Waves, Beaches, Wave-Structure Interactions, Elsevier,
Amsterdam - New York - Tokyo.
SILVESTER, R., 1986, Use of grout-filled sausages in coastal structures, ASCE, J. Waterway, Port
Coastal and Ocean Engineering, Vol. 112, No. 1, January.
SILVESTER, R., 1990, Flexible Membrane Units for Breakwaters, in 'Handbook of Coastal and Ocean
Engineering, John B. Herbich, ed., Vol. 1, pp. 921-938.
SPRAGUE, C.J., 1994, Dredged material filled geotextile containers, prepared for the US Army Corps of
Engineers, Vicksburg, MS.
STOUTJESDIJK, T, 1996, Geotechnical aspects of alternative systems, Delft Geotechnics.
TANAKA, M., T. OYAMA, T. KIYOKAWA, T. UDA and A. OMATA, 1990. Wave control by flexible
underwater mound, Proc. Offshore Tech. Conf., OTC 6405.
TETRA TECH, 1982, Longard Tube Applications Manual (Aldek A-S Longard), U.S.A.
TUTUARIMA, W.H. and W. van WIJK, 1984, ProFix mattresses - an alternative erosion control system, in
Flexible Armoured Revetments incorporating geotextiles, T. Telford.
VAN OORD ACZ BV., 1995, Test programme of geocontainers, Van Oord ACZ/Nicolon.
Van der HIDDE, 1995, BEROSIN, product info., Bureau van der Hidde, Harlingen, P.o.B. 299, NL.
Van der WAL, M., 1997, Hydraulic aspects of the dimensioning of geosystems, Delft Hydraulics,
Report H 3214.
WOUTERS, J., 1995, Stability of geosystems (in Dutch), Delft Hydraulics, Report H1930.

25