Вы находитесь на странице: 1из 51

Geosynthetics applications for the mitigation of

natural disasters and for environmental protection


H. Brandl
Emeritus Professor; Institute for Geotechnical Engineering, University of Technology, 1040 Vienna,
Austria. Telephone: +431 58801 22101, Telefax: +431 58801 22199, E-mail: heinz.brandl@tuwien.ac.at
Received 7 July 2011, revised 17 August 2011, accepted 17 August 2011
ABSTRACT: The paper first describes the versatile application of geosynthetics for the mitigation
of floods, landslides, rockfalls, debris flows and avalanches. It focuses on dykes or flood protection
dams, on geosynthetic-reinforced stabilising fills (up to 130 m height) and barrier dams.
Geosynthetic-reinforced floating embankments (up to 70 m height) in creeping slopes and seismic
areas show clear advantages over rigid structures (e.g. bridges) not only from a geotechnical point
of view but also regarding economy, maintenance and environmental aspects. Environmental
protection is predominantly considered by gaining renewable energy from the ground via energy-
geosynthetics. Several other applications are also mentioned. Compaction optimisation and control
of geosyntheticsoil structures is recommended by roller-integrated CCC (continuous compaction
control), thus improving their behaviour significantly.
KEYWORDS: Geosynthetics, Natural disaster mitigation, Flood protection, Rockfall protection,
Landslide protection, Environmental protection, Floating embankments, Energy-geocomposites
REFERENCE: Brandl, H. (2011). Geosynthetics applications for the mitigation of natural disasters and
for environmental protection. Geosynthetics International, 18, No. 6, 340390. [http://dx.doi.org/
10.1680/gein.2011.18.6.340]
1. FLOOD PROTECTION
1.1. General
Climate change requires a close cooperation between
hydrology, hydro engineering and geotechnical engineer-
ing. The increasing frequency and magnitude of floods
indicate that the previous prognoses of such events should
be revised for many regions. Consequently, the improve-
ment of old dykes or flood protection dams and the
construction of new ones has become essential since the
1990s. Environmentally optimal solutions are achieved by
(re-)creating retention areas, combined with dams that can
be overflowed, thus flattening the flood wave. Besides
these permanent flood protection measures, temporary
ones are increasingly required for emergency situations.
Geosynthetics have also proved very suitable for mobile
systems.
The topic of coastal protection should also be men-
tioned in the list of possible applications of geosynthetics
in the field of natural disaster mitigation. However, it will
not be discussed in this 2010 Giroud Lecture, because it
has already been dealt with in the 2006 Giroud Lecture
delivered by C. Lawson (Lawson 2008).
Quality assessment of existing structures, design aspects
of new ones, and the required safety factors depend on
purpose, risk potential and monitoring intensity. Therefore
dykes (permanently loaded by water) and flood protection
dams (only temporarily loaded by water) should be distin-
guished. Most geotechnical aspects, however, are relevant
for both.
1.2. Future trends and residual risks
Floods have affected millions of people worldwide in
recent decades. In several regions the magnitude and
frequency of flood waves have increased dramatically
since long-term measurements and historical reports have
existed. Figure 1 shows an example from Austria, where a
2000 to 10 000-year flood event was back-calculated from
the flood disaster in the year 2002. Such hitherto singular
values cannot be taken as design values for flood protec-
tion dams, but they underline the need for local overflow
crests or spillway sections. Moreover, they clearly demon-
strate that a residual risk is inevitable despite most
costly protective measures. The so called absolute safety
as frequently demanded by the public, by politicians, by
the media or by jurists cannot be achieved in reality.
1.3. Failure modes of dykes and dams
Knowledge of possible failure modes is an essential
prerequisite both for a reliable quality assessment of
Geosynthetics International, 2011, 18, No. 6
340 1072-6349 # 2011 Thomas Telford Ltd
existing dykes and dams and for an optimised design of
new ones, in connection with the application of geosyn-
thetics. Figure 2 summarises the dominating failure modes
for typical ground conditions along rivers (near-surface,
low-permeability sandy to clayey silts underlain by high
permeability sand or gravel)
slope failure due to excessive pore-water pressures,
seepage or inner erosion
overtopping of the dyke/dam crest
slope failure due to a rapid drop of the flood water
level
hydraulic fracture
surface erosion and failure of the water-side slope
due to wave action
piping due to animal activities, especially from
beavers and rats
unsuitable planting of dykes (especially trees with
flat roots).
Hydraulic failure is frequently underestimated, but may
occur in different forms (e.g. Eurocode 7; CEN 2004).
By uplift (buoyancy). The pore-water pressure under
the low-permeability soil layer exceeds the over-
burden pressure.
By heave. Upward seepage forces act against the
weight of the soil, reducing the vertical effective
stress to zero; soil particles are then lifted away by
the vertical water flow. This boiling dominates in
silty-sandy soil, and is combined with internal
erosion (Figures 3a, 3b and 4).
By internal erosion. Soil particles are transported
within a soil stratum or at the interface of soil strata
(Figures 5 and 6). This may finally result in
regressive erosion, leading to ground failure of the
dyke or dam.
By piping. Failure by piping is a particular form of
internal erosion, where erosion begins at the surface,
and then regresses until a pipe-shaped discharge
tunnel is formed (Figure 7). Failure occurs as soon as
the water-side end of the eroded tunnel reaches the
river bed or bottom of the reservoir. Frequently,
several tunnels develop (Figure 8). This process may
be induced or significantly promoted by animal
activities, as field observations over many years have
revealed (Figure 2).
Hydraulic failure may reach several tens of metres away
from dykes or dams, as experience has shown. This could
be observed even for low flood protection embankments
with a relatively small hydraulic gradient.
1 10 100 1000 10000
x-year flood event
M
a
x
i
m
u
m

d
i
s
c
h
a
r
g
e
:

m
/
s
3
0
100
200
300
400
500
Flood 2002: ca. 420 m /s
Back calculated: 2000 to 10000-year event
3
Hitherto observed event
GEV-LM Extrapolation
Figure 1. Statistics for the annual maximum discharge values
of the river Kamp in Austria (Gutknecht et al. 2002)
Water outflow
Eroded material
Cracks
Low-permeability soil layer
Low-permeability soil layer
Low-permeability soil layer
Low-permeability soil layer
Low-permeability soil layer
Low-permeability soil layer
Low-permeability soil layer
Low-permeability soil layer
Sand
Sand
Sand
Sand
Sand
Sand
Sand
Sand
Figure 2. Failure modes of dykes and dams: near-surface
stratum of low permeability (possibly with local windows).
After VDZ (2002)
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 341
Geosynthetics International, 2011, 18, No. 6
Eurocode 7 (CEN 2004) states that in situations where
the pore-water pressure is hydrostatic (negligible hydrau-
lic gradient) it is not necessary to check other than for
failure by uplift. In the case of danger of material
transport by internal erosion, filter criteria should be
used. If the filter criteria are not satisfied, it should be
verified that the critical hydraulic gradient is well below
the design value of the gradient at which soil particles
begin to move.
Experience has shown that the magnitude of the critical
hydraulic gradient where internal erosion begins is fre-
quently overestimated. Figure 9 summarises the critical
values on the basis of field observations, geotechnical
measurements, literature and long-term experience for
different soils. For comparison, the conventional criterion
(i
crit
9/
w
), Lanes criterion, and the critical zones after
Eurocode 7 (CEN 2004) or Chugaev (1965) respectively
are also plotted in the diagram.
Filter protection is generally provided by the use of
non-cohesive granular material (natural soil) that fulfils
adequate design criteria for filter materials. Filter geotex-
tiles have been used increasingly since the early 1970s.
Common filter criteria for soils are from Terzaghi and
Sherard, and for geotextiles from Giroud (2003, 2010) and
Heibaum et al. (2006). All criteria have particular limit-
ations, whereby non-cohesive and cohesive soils have to
be distinguished. Whereas two criteria are sufficient for
granular filters (the permeability criterion and the reten-
tion criterion), four criteria are required for geotextile
(a)
(b)
Figure 3. Boiling behind dykes indicates inner erosion and
beginning of hydraulic failure: (a) randomly distributed
multiple boiling; (b) detail of an erosion spot
Figure 4. Boiling volcanos after the flood
Phase 1 Phase 1 Phase 2 Phase 2
(a) (b) (c)
Figure 5. Hydraulic fracture of dykes or flood protection
dams due to seepage through or beneath the dyke or dam
(Ziems 1967): (a) suffusion (fine particles move into pore
voids of coarse grain fractions); (b) contact erosion at the
interface of soil strata; (c) internal erosion in steady-state
flow condition
Figure 6. Contact erosion perpendicular or parallel to an
interface of soil strata
342 Brandl
Geosynthetics International, 2011, 18, No. 6
filters (Giroud 2010): the porosity criterion and the
thickness criterion also have to be considered.
1.4. Quality assessment of dykes and dams
Comprehensive in situ investigations have disclosed that
the quality of dykes and dams with a height, H , 15 m is
mostly lower than that of higher earth structures. There-
fore low but long flood protection dams may involve
rather high risk potential. Dams of H > 15 m, however,
have already been constructed for decades according to
the strict regulations of ICOLD.
When assessing the quality of dykes or flood protection
dams, the purpose and risk potential of the structure
should be taken into consideration. Dykes as fill dams
directly along reservoirs, rivers or canals are continuously
1
6
3
4
1 Free water table
2 Piezometric level in permeable subsoil
3 Low-permeability soil
4 Permeable subsoil
5 Possible well; starting point for pipe
6 Possible pipe
2
1
5
Figure 7. Example of conditions that may cause piping (Eurocode 7; CEN 2004)
Figure 8. Piping through a railway embankment during an
already sinking flood
Medium dense stiff
i
c
r
i
t
Fine Fine Fine Medium Medium Medium Coarse Coarse Coarse
Clay
Silt Sand Gravel
Cobbles Boulders
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0

G,dst

G,dst
D
A
E
Hydraulic fracture EN1997-1
i
crit

W
0.001 0.002 0.006 0.02 0.063 0.2 0.63 2 6.3 20 64 100 200
Grain size (mm)
B
C
Figure 9. Critical hydraulic gradients for hydraulic fracture (internal erosion) (Brandl and Hofmann 2006); i
crit:
depends not
only on grain size distribution and density/stiffness but also on flow pressure;
G,dst
partial safety factor for permanent
unfavourable effects
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 343
Geosynthetics International, 2011, 18, No. 6
monitored, whereas flood protection dams are loaded by
water only during floods. The latter are therefore more
critical, as experience has shown.
Besides levelling of the crest and control of the
geometric data, the following (geotechnical) investigations
have proven successful
seasonal field observations, especially during floods
documentation of water discharge and localisation of
wet slope spots and zones during floods, depending
on the magnitude and duration of the flood wave
aerial photographs, especially during and after floods
(and in normal periods for comparison)
spot-checking by soundings, borings or exploratory
pits
field tests and laboratory tests
geophysical investigations
tracer testing
infrared photographs for localisation of seepage
roller-integrated continuous compaction control
(CCC).
Investigations in steps provide the best results, and are
the most cost-effective. Conventional spot-checking gives
only random data, whereas investigations covering the
entire area provide more detailed information. Conse-
quently, geophysical methods have been increasingly used
over the last 10 years. However, experience has shown that
a reliable interpretation is possible only if at least two or
three different methods are applied (e.g. geoelectrics,
micro-gravimetry, combined surface wave/refraction seis-
mic). Geophysical methods and electric tracers are also
used to check the effectiveness of sealing elements (sur-
face sealings, core sealings, cut-off walls). Multi-sensor
survey systems together with spatially targeted electrical
tracers make it possible to localise leakage areas, and
hence leaks. Moreover, they have proved suitable for
continuous leakage monitoring of sealing systems (liners,
cut-off walls).
A complete quality assessment of the dyke or dam crest
is made possible by roller-integrated continuous compac-
tion control (CCC). This method registers the interaction
between roller and ground (for details see Section 5). The
measuring depth reaches 2 to 2.5 m, and thus covers the
most critical zone of dykes and dams. Weak points are
easily localised, which is especially important for struc-
tures with crests that can be overflowed (spillway section).
However, if the crest has a top cover of concrete, asphalt
or riprap, or if it is very uneven, CCC cannot be used.
The advantage of CCC is not only full control but also
a compaction of weak zones. Figure 10 shows some
results along an old dyke that was severely attacked by
the previous flood. The weak points and zones were not
visible along the crest, but could be clearly localised by
CCC during the first roller pass. After six roller passes
the quality of the top 2 m of the dam had improved
significantly, although relatively weak spots still existed.
However, the control values exceeded the lower limit
value (30 for this particular fill material) with only one
exception (23).
1. Roller pass
6. Roller pass
120
120
100
100
80
80
60
60
40
40
20
20
0
0
110/100
110/100
110/150
110/150
Cursor: 110/190
Required limit
Cursor: 110190
Required limit
15.9
30
29.1
30
Mean value: 38.6
SD: 20.5
Mean value: 56.3
SD: 18.2
Max: 89.1
Min: 10.3
Max: 94.7
Min: 23.0
110/250
110/250
110/300
110/300
110/200
110/200
Pt1: HMV 77.8
Pt. 2: HMV 42.7
Pt.3: HMV 29.5
Pt.4: HMV 23.0
Figure 10. Results from roller-integrated continuous compaction control (CCC) of an old flood protection dam after a severe
flood. Weak spots are clearly visible; also improvement after six roller passes. Compaction degree/stiffness (dimensionless value)
is plotted against the chainage of the dam
344 Brandl
Geosynthetics International, 2011, 18, No. 6
1.5. Geosynthetics application for dykes and dams
Most dykes and flood protection dams are lower than 10
to 15 m, but are clearly longer than dams for hydropower
generation. Moreover, the latter are cross structures
whereas dykes and flood protection dams usually represent
longitudinal barriers. Consequently, different design rules,
maintenance and monitoring aspects have to be consid-
ered. This leads to significantly increased application of
geosynthetics for dykes and flood protection dams.
Geosynthetics have proved successful for emergency
and temporary measures as well as for permanent pur-
poses, and for contingency plans (successful defence of
dykes and dams against severe floods). Furthermore,
geosynthetic sensor mats serve for monitoring of critical
zones.
1.5.1. Emergency or temporary measures
Placing filter stable counterweights (berms) can prevent
hydraulic failure of the dyke or dam by seepage or uplift,
or by internal erosion and piping. Sandbags are preferred
for local stabilisation, and sheets of filter geotextiles
(covered with sand, gravel, or other granular material) for
larger critical zones (Figures 11 and 12). Water outflow
must not be prevented, as it would create excessive pore-
water pressures and favour sudden failure. Consequently,
placing impermeable geomembranes or plugging weak
spots with clay at the land side is counterproductive
(although it is in many cases the instinctive reaction of
those defending the dyke or dam).
If the dyke or dam is already broken, progressive failure
can be avoided by using a helicopter to place, first, stiff,
heavy elements (e.g. military anti-tank steel crosses) to
form a skeleton, and then fill large bags and finally
smaller sandbags (Figure 13). After the flood, and removal
of the emergency elements, the waste consisting of soil
and geosynthetics should be recycled (Figure 14).
Mobile flood protection systems may be stiff panels
inserted in fixed toe elements and other modular systems
of metal, reinforced concrete or synthetic material.
Figure 11. Hydraulic failure prevention by covering local
water outflow with permeable sandbags (small geotextile
bags)
Figure 12. Hydraulic failure prevention (emergency measure)
by covering the entire dam slope and berm with gravel
placed on a filter geotextile
Figure 13. Large bags and sandbags between anti-tank steel
crosses to close a broken dyke section (emergency measure).
Sequence of placing: 1, steel crosses; 2, large bags; 3.
sandbags
Figure 14. Soilgeosythetics waste after the flood when
removing the emergency elements
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 345
Geosynthetics International, 2011, 18, No. 6
Geosynthetics are a promising alternative whereby semi-
permeable and impermeable systems have proved success-
ful. Geotextile container solutions show a great versatility:
small bags, large bags, containers, tubes and mattresses.
Many of them are also used for permanent purposes
(Saathoff et al. 2007).
Pre-filled sandbags, which can be easily carried and
placed by one person, are frequently preferred to larger
geotextile sand containers. They have already been suc-
cessfully used for decades. However, the better the access
and placement conditions for technical equipment, the
larger the containers that can be used. Finally, if there is
no access to critical zones on land, large bags can be
placed by helicopter.
In all cases the geotextile should have high robustness
(puncture resistance), high elongation behaviour and inter-
face friction. Flexible behaviour allows the geotextile
container to mould itself in with the existing features, and
also allows a certain degree of self-healing of the
structure. Unavoidable damage from driftwood and similar
objects is also then minimised.
The permeability of structures made of geotextile sand
containers depends primarily on the size of the container
elements and the method of placement. The grain size
distribution of the infill is negligible. Recio and Oumeraci
(2008) showed that the flow through the structure is
governed solely by the gaps between neighbouring con-
tainers, and that the flow through the sand fill in the
containers can be neglected. Figure 15 compares two
models (out of 11) providing the worst and best results.
The permeability coefficients obtained are in the order of
5 3 10
2
m/s and 7 3 10
3
m/s, respectively, despite a
rather similar width of the structures. Furthermore, place-
ment of the containers such that the contact areas between
containers were maximized resulted in the highest stability
against wave action.
Figure 16 shows a modular system of metallic grid-
boxes lined with nonwoven geotextiles and filled with
sand. Such elements are preferably used along roads and
in settlements and cities. By placing them both side by
side and on top of each other, massive and high barrier
structures can be constructed. Water discharge is largely
negligible, roughly comparable to that through a wall of
closely placed geotextile sandbags.
Synthetic impermeable tubes (usually of PVC) are easy
to transport. They are connected by straps, then inflated
with air, moved to a fixed position and filled with water
(Figure 17). Geotextile tubes to mitigate the wave-
breaking process in meandering rivers or to fill broken
sections of dams are less used, but they have proved very
suitable as low-crested submerged structures to reduce the
incident wave energy on shorelines.
Mobile flood protection systems require good and
repeated training of the emergency staff to assemble and
activate the barriers as quickly and effectively as possible.
Medium
Cross-section Front view Plan view
GSC, one above the
other
Containers above each other,
maximal size of gaps
Large
Large
Large
M
o
d
e
l

1
M
o
d
e
l

1
1
GSC, overlapped
Medium
Medium
Medium
M
e
d
i
u
m
Overlapped
Overlapped
Gap
Figure 15. Models to determine the hydraulic permeability of structures made of geotextile sand containers. Significant
influence of container size and method of placement on hydraulic permeability: k 5 3 10
2
m/s (Model 1), k 7 3 10
3
m/s
(Model 11) (Recio and Oumeraci 2008)
Figure 16. Mobile flood protection system with gabions of
lined nonwoven geotextiles filled with sand. The structure
finally withstood a 100-year flooding event
346 Brandl
Geosynthetics International, 2011, 18, No. 6
1.5.2. Permanent measures for dykes and flood protec-
tion dams
Figure 18 gives a schematic selection of geosynthetics
application for dykes and flood protection dams. Not
included are geosynthetics for overflow (spillway) sec-
tions, for vertical cut-off walls, for geotextile sand con-
tainers and tubes. The latter are used for erosion control,
scour protection, scour fill-in, groynes and breakwaters.
Experience gained in coastal engineering can be widely
adapted to rivers. There are, of course, certain differences
between constructing new dykes and flood protection
dams and repairing or refurbishing existing structures, but
the basic design requirements are largely the same.
Dykes or dams fully reinforced with geosynthetics, as
indicated in Figure 18, are rather exceptional cases, for
instance along the outer bank of a river, or at a local
narrowing of the cross-sectional flow. Transition zones
from rigid structures (e. g. culverts) to the flexible dam
fill and overflow sections may also exhibit geosynthetic
reinforcement.
Figure 19 shows versatile application of geosynthetics
for the rehabilitation of a section of the dyke at the river
Elbe after the great flood in 2002. Slope sealing at the
water side of the dyke was widely used there, because
geomembranes or geosynthetic clay liners can be placed
very quickly.
River bank and canal slope protection also has an
influence on possible flood damage. Solutions with geo-
synthetics generally have lower construction and lifetime
costs than rigid structures.
To sum up, geosynthetics serve the following purposes
for dykes and flood protection dams
horizontal, vertical and inclined filters and separation
elements
dyke slope sealing at the water side
reinforcement of fill, crest zone, access road (for
dyke defence)
surface erosion protection
vertical cut-off walls
protection against voles and beavers.
Inclined clay cores of flood protection dams commonly
Figure 17. Mobile flood protection system with synthetic
tubes
Dyke/dam
defence way
Dyke/dam
defence way
Drainage
gravel
Drainage
gravel
Separation geotextile
Filter geotextile
Separation geotextile
Filter geotextile
Separation geotextile
Filter geotextile
Separation geotextile
Filter geotextile
Erosion protection mats
Erosion protection mats
Erosion protection mats
High level
Rip-rap
Low level
Rock fill
or gravel
Excavation
in steps
Gradient
4
5
1
5
3
4
5
2
6
Vegetation soil cover
Flood
Geomembrane
Geosynthetic clay liner
Geosynthetic reinforcement
Sand
Figure 18. Application of geosynthetics for dykes and flood protection dams. Schematic examples, not to scale. 1, filter
geotextile for drainage fill; 2, geosynthetic-reinforced dyke/dam; 3, filter geosynthetics below rockfill or riprap; 4, geosynthetics
for the access way for dyke/dam defence during floods; 5, geosynthetics for erosion protection; 6, geomembrane or geosynthetic
clay liner
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 347
Geosynthetics International, 2011, 18, No. 6
lead to a lower slope stability than vertical cores or cut-off
walls. This refers especially to a quick drop of the water
level, causing superficial sliding of the water-side slope,
leaving the clay core unprotected against erosion. In the
case of geomembranes or geosynthetic clay liners, near-
surface slips are also critical, but can be repaired more
easily than along softened, partially eroded clay.
Hydraulic failure may be prevented mainly by two
measures on the land side of a dyke or flood protection
dam
installing trenches or relief columns or drainage
wells
filling of berms, thus displacing the possible starting
point of inner erosion or piping further away from
the structure, and decreasing the hydraulic gradient
at this point. Such berms should be constructed as
access roads for quick and easy dam defence in the
case of severe floods.
In many cases berms merely move the hydraulic
problem further away from the dyke or dam, and retro-
gressive inner erosion may finally reach it in the long term
(after several floods). Boiling and internal erosion have
been observed up to 20 to 50 m away from dykes and
dams, even though they were only 3 to 6 m high. More-
over, wide berms are frequently not possible under
confined space conditions: therefore drainage trenches are
preferred in these circumstances. However, trenches ex-
cavated in very soft soil collapse immediately before
geotextiles and fill material can be placed. The installation
of trussed retaining panels would be too expensive. These
problems could be overcome by developing relief granu-
lar columns, jacketed with a filter geotextile:
Jacketed (coated) stone or gravel columns have been
installed in Austria since 1992. At first they were used
mainly for drainage purposes, for instance as drainage
walls to improve the stability of old flood protection earth
dams. This method has significant construction advantages
over conventional drainage trenches in loose or soft soil.
In critical cases the coated columns are combined with
other measures for dam refurbishment (Figure 20). The
drainage material (usually clean 4/32 mm, 8/32 mm or 16/
32 mm grain) is lowered by vibroflotation, whereby the
vibrator is wrapped with a nonwoven geotextile (tied
together at the toe of the vibrator).
The conventional top-feed process of the vibro tech-
nique is not suitable for jacketed granular columns. In this
case the sophisticated vibroflotation technique with
bottom-feed vibrators is required. The main advantage of
this method is that the vibrator remains in the ground
during installation, making the technique ideal for un-
stable ground and high groundwater levels. The granular
material is discharged from skips into the chamber at the
top of the vibrator and placed at depth (Figure 21). To
avoid geotextile damage, the vibrator is sometimes first
lowered without the geotextile sleeve into the ground to
displace soil. This has proved suitable in coarse or stiff
subsoil, but is not necessary in fine-grained soft ground.
A leak in the geotextile is quickly recognised, owing to an
increasing volume of fill material during vibration.
Figure 22 shows an alternative technique for relatively
short gravel columns (up to 6 m). Column excavation is
performed with continuous-flight auger-piling equipment.
Then a steel tube with a slightly greased surface and
covered with a filter geotextile is lowered into the ground.
When being withdrawn the pipe is filled with gravel,
whereby the geotextile remains wrapped in the ground,
thus enveloping the gravel.
Commonly, mechanically bonded continuous filament
nonwoven geotextiles of polypropylene are used. In the
Air side
80
75
70
Stone cover ( 60 cm)
on filter nonwoven
d
Sandgravel
wrapped into filter nonwoven
Geogrid
Dyke defence path
Old dyke
Dyke crest
on filter nonwoven
V
:H
1
:3

V
:H
1
:2

Water side
Vegetation soil ( 25 cm)
Cover soil ( 75 cm)
Geosynthetic clay liner (GCL)
Sub-base (old dyke section)
d
d

Design water level


0 5 10 15 20 25 30 35 40 5
Figure 19. Dyke remediation in a section along the river Elbe (Heerten 2006)
Thin diaphragm wall
Drainage pipe
on nonwoven
geotextile
Sleeve reinforced
stone columns
(a 1.25 m)
Compaction by
vibroflotation
(Silty) sand
gravel
Clay, silt
Sand, gravel
Figure 20. Improvement of static and hydraulic stability of
an old flood protection dam
348 Brandl
Geosynthetics International, 2011, 18, No. 6
case of large construction sites, the coating is already
prefabricated by the manufacturer and distributed in a
tubular shape with a needle-sewn seam fitting specifically
to the vibrator of the vibroflotation equipment. The
geotextile characteristics listed in Table 1 have proved
successful.
To optimise the design of the position, spacing, depth,
diameter and hydraulic capacity of geotextile jacketed
relief gravel columns requires numerical modelling for
three-dimensional unsteady flow conditions. Multi-layered
soil systems always exhibit permeability coefficients that
are different in the horizontal and vertical direction.
Therefore numerical models and calculations should be
calibrated by site measurements or observations along
previous projects, or by field testing.
The tops of relief columns should be covered with
coarse drainage material, wrapped in filter geotextiles
(Figures 23 and 24) for longitudinal or transverse drain-
age. This drainage layer should carry an access road for
easy dam defence in the case of severe floods.
Figure 24 shows a cross-section through a new flood
protection dam after removal of the old one, which had
been destroyed by a severe flood. The coated gravel
columns (diameter 0.7 m) usually exhibit a spacing be-
Figure 21. Installation of geotextile-jacketed drainage
columns (relief gravel columns) by bottom-feed vibroflotation
Figure 22. Installation of geotextile-jacketed drainage
columns by inserting a sleeved steel pipe into a pre-bored pile
excavation. After pipe withdrawal, the gravel-filled geotextile
hose remains in the ground
Table 1. Geotextile characteristics for jacketed drainage
columns
Property Value
CBR puncture resistance (kN) 3.85
Strip tensile strength (kN/m) 24/24
Elongation at maximum load (%) 80/40
Cone drop test (hole diameter) (mm) 15
Mass (g/m
2
) 325
Thickness at 2 kN/m
2
(mm) 2.5
Figure 23. Top of geotextile-jacketed drainage columns
embedded in coarse drainage layer as sub-base of the access
road for quick dam defence in case of severe floods
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 349
Geosynthetics International, 2011, 18, No. 6
tween 1.5 and 7.0 m, depending on local factors (geo-
technical and ecological parameters, infrastructure, risk
potential etc.); spacing is commonly about 4 m. The
water-side dam slope is covered by a net for protection
against beavers (see Section 7).
In many regions, the crest levels of dykes and flood
protection dams will need to be raised in the future.
However, because of climate change, the prognoses in-
volve several uncertainties. Therefore these structures
should be adapted to allow overflow in spillway sections,
or at least wave overtopping. This minimises random
failures of dykes and dam. The results of comprehensive
field tests on sea defences subject to wave overtopping
(Akkerman et al. 2007) can also be applied for dykes and
dams along rivers. Three-dimensional reinforcement grids
with additional soil erosion protection represent a flexible
alternative to riprap or even stiffer structures (Figure 25).
1.5.3. Prevention of seepage
From strict theory, seepage through dykes or flood protec-
tion dams cannot be prevented, but in practice it can be
limited to nearly zero (considering evaporation etc.). This
can be achieved by
homogeneous dykes or dams of low-permeability
fill material
zoned dykes or dams with clay cores (vertical in the
centre, inclined at the water side)
cut-off walls (independent of the fill material).
Geosynthetics have proved suitable as an alternative to
clay cores and diaphragm cut-off walls (slurry trench
walls). Geosynthetic cut-offs are inserted into the dyke or
dam by vibration, or by lowering them into a deep trench
that is then backfilled (Figure 26). Leak detection is
possible by checking the hydraulic potential on either
side of the sealing element, by thermometry, and by
geophysical methods (mainly geoelectrics). Moreover,
innovative systems with integrated leak detectors are
available.
Slurry trench walls or deep-mixing walls, however,
have a higher resistance against beaver attack than
geosynthetic screens; they have a statical function
(although theoretically neglected for thin diaphragm
walls) for example, where not only a sealing function
is required but also the ability to take higher forces for
special cut-off walls; and they may reach significantly
deeper into the natural ground to reduce or prevent
seepage beneath the dyke or dam. Furthermore, they have
proved suitable for defined fracture sections of dams: In
an emergency it may be necessary to open the dam at
specific points in order to avoid irregular failures at
random zones. A domino-like progressive failure extend-
ing from the designed fracture section must be prevented.
Consequently, sheet pile elements are also used for such
structures.
HW100
Low-permeability stratum
(with windows)
High-permeability stratum
(aquifer)
Crest way
T
h
i
n

d
i
a
p
h
r
a
g
m

w
a
l
l
Potential line with relief
Defence way (asphalt way)
Relief trenches
or columns
Figure 24. Standard cross-section of a new 75 km long flood protection dam in Austria
Figure 25. Stiff crest structure of a wave overtopping and
spillway section of a dyke damaged by flood
Figure 26. Backfilled geomembrane cut-off wall for low flood
protection dams
350 Brandl
Geosynthetics International, 2011, 18, No. 6
1.5.4. Mitigation of flood damage
In Austria nearly 3000 flood protection projects are under
construction, design and pre-planning until the year 2015.
This requires multidisciplinary cooperation, and also con-
sideration of infrastructure, of environmental protection,
and of local, public and legal aspects. Plans for dam
maintenance, precautionary measures and emergency ac-
tions have been developed since the flood disasters of
2002 and 2006. Contingency plans (based on risk ana-
lyses) comprise scenarios from slight but frequent floods
to worst-case events (maximum credible water levels).
Another essential prerequisite for successful mitigation
of flood effects is the comprehensive education and
training of task forces comprising authorities, organiza-
tions, professional groups and volunteers. Flood wave
prognoses for dam overtopping or failure, contour maps of
water level, and warning, alarm and evacuation plans are
already standard, and are highly appreciated by the
Austrian public.
Each new or refurbished dyke or flood protection dam
will be equipped with a parallel flood defence road that
makes quick access possible, so that necessary mitigating
measures can be taken without delay. Materials for such
measures (mainly filter geotextiles) are stored in the
vicinity.
These concepts correspond to a judgment by the
European Court for Human Rights stating the states
responsibility to protect its citizens from natural disasters
to an appropriate (not an absolute) extent.
2. GEOSYNTHETIC-REINFORCED
BARRIERS AGAINST ROCKFALL:
MODEL TESTS
2.1. Measures against rockfall
In mountainous regions, large-scale rockfalls have become
an essential element in regional planning (Figure 27).
Well-known disasters have drawn the attention of both the
public and the authorities to the need for extensive meas-
ures to protect critical areas. Optimised investment
addresses several questions concerning risk analysis, such
as the probability of a severe event and, particularly, the
amount of damage it causes. Taking economical consid-
erations into account leads to a limited risk reduction in
many cases of rockfall protection.
In recent years the prediction of rockfall has undergone
significant improvement regarding the trajectories of the
falling blocks by the use of highly developed computer
modelling, which considers such factors as mass, geome-
try, damping and vegetation. Ultimately this yields the
blocks kinetic energy at any position. In most cases,
though, prediction of the behaviour of an unstable block
itself involves uncertainties, and therefore often requires a
wide parameter variation in calculation and design.
The development of structures to guard against rockfall
has received an enormous boost over the last 15 years in
the field of netting, where the cost benefit ratio can be
noticeably improved by the high absorption of deforma-
tion energy. These results were also facilitated by full-
scale fall tests, in which not only were further types of
netting developed, but also new forms of evaluating the
absorbed forces were applied.
Descoeudres (1997) cites various structures against
rockfall, rated according to their deformation energy
(Figure 28). Among these passive measures against rock-
fall, earthfill barriers and especially reinforced embank-
ment dams (barriers) take up the prime position. In
contrast to flexible netting, the energy-absorbing effect of
protective barriers is provided mainly by their mass.
Sufficient space, and a topography that makes the rock-
falls transit area sufficiently well known, are prerequi-
sites for the choice of an embankment barrier.
Furthermore, the risk and the quantity of the expected
rockfall have to justify the costly option of constructing an
earthfill barrier rather than using netting.
As the different trajectories of the falling blocks and
their rotational behaviour can be predicted to a high
degree by computer modelling, it seems appropriate to
take this into account not only when positioning the
barrier but also in the design of its upper (mountain-
facing) slope. Obviously, geosynthetic-reinforced slopes
offer more possibilities in shaping the embankment
properly to prevent the blocks from running over the
barrier. Additionally, the catch basin can be enlarged by
steeply sloped barriers. Questions on the comparability of
barriers with regard to their overall resistance against
dynamic impact seemed so far to have remained unan-
swered, and the economical aspects of such structures
have not been addressed.
Full-scale tests commonly require a large effort when
comparing various kinds of earthfill barrier under specific
conditions. Therefore, qualitative and semi-quantitative
model tests have been preferred for parametric studies, and
to investigate the interacting factors of the geosynthetics
arrangement, anchoring lengths, and degree of compaction.
2.2. Test set-ups and performance
Walz (1982) describes qualitative model tests in soil
mechanics primarily as a method to recognize the failure
mechanism of stability problems. Starting from this phi-
Figure 27. Severe rockfall along an expressway; cars were
crushed
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 351
Geosynthetics International, 2011, 18, No. 6
losophy, a series of 20 dynamic 1g model tests were
carried out on protective barriers against rockfall, scaled
at 1:50 (Blovsky 2002). In the soil mechanics laboratory
at the Vienna University of Technology special attention
was directed to the measurement of forces, acceleration
and deformations in order to gain comparable results, and
to enable systematic parametric studies.
In order to record the dynamic impact exactly, the
rockfall was simulated by a rigid pendulum that contained
dynamic force and acceleration transducers. For exact
measurements it was necessary to retain the six degrees of
freedom that a single falling block has. Interpretation of
the measured data was supported by deformation gauges
in the embankment, and optical recording by two digital
video cameras. Figure 29 shows the pendulum with its
hemispherical penetration surface, dynamic force and
acceleration transducers, and a threaded pole for addi-
tional weight.
The models were constructed in a steel frame structure
with formwork boards as side walls for the cross-sections
of the barriers, to serve as parallel guides. To achieve a
uniform density, the model soil (sand with a low degree of
uniformity) was compacted in thin, laterally boarded
layers in such a way that the required soil mass could be
weighed and controlled (Figure 30a). Without the form-
work for the embankment the conditions in the impact
area would not have been reproducible.
To achieve a maximum slope inclination of 3:1 in the
Grates
Diagonal ringnets
Anchored nets
Can
Rock sheds
Earth fill
Reinforced barriers
10 20 50 100 200 500 1000 2000 5000 10000 50000
kJ
Figure 28. Structures against rockfall according to their deformation energy (Descoeudres 1997)
Acceleration
transducer
Thread pole
for additional
weight
Force
transducer
Hemispherical penetration
surface (wooden)
Figure 29. Pendulum simulating the rockfall
(a)
(b)
Figure 30. Construction of the standard model for
simulating rockfalls on barrier fill dams
352 Brandl
Geosynthetics International, 2011, 18, No. 6
models, the soil had to be compacted at Proctor water
content, thus gaining a certain amount of apparent cohe-
sion, which of course would not exist within a non-
reinforced barrier prototype for a long time. These steep
slopes were carried out to compare non-reinforced and
reinforced barriers.
As already mentioned, the dynamic impact was simu-
lated by a rigid pendulum to provide an exact placement
of the transducers. The pendulum was orientated horizon-
tally in half of the barrier height at the moment of impact.
To gain information about the behaviour of the barriers
under that kind of load, each test contained several strokes
(rockfalls), starting with a small impulse and increasing
to a maximum impulse to destroy the model. The loading
history of each test had to be exactly the same, to achieve
full comparability. To control the increasing impulse,
additional weight and/or the release height were varied
(Figure 30b). The release of the pendulum from its set
height was effected by a steel wire and a special clamp
that provided release almost without jerking. This method
was essential, since triggering of the measurement was
executed via the acceleration signal of the pendulum in
order to gain sufficient data from the whole period of
impact at a maximum measuring rate. Optimising these
parameters led to a possible maximum measuring time of
2.5 s at 2400 Hz dynamic rate.
2.3. Parametric studies
The main goal of the qualitative model tests was to
compare the influences of geometry and compaction on
the one hand, and the effectiveness of geosynthetic rein-
forcement on the other hand. Therefore 20 model tests
were arranged as shown in Figure 31 (non-reinforced) and
Figure 32 (reinforced). Each group started from so-called
standard tests (Nos 01 to 03 non-reinforced, and No. 10
reinforced), in which both slopes were inclined at 4:5, and
the compaction was 100% of standard Proctor density.
Subsequently, individual parameters such as slope inclina-
tion, degree of compaction and the arrangement of the
reinforcement were varied. Combinations of altered para-
meters were considered in this first series of model tests.
The first three tests were conducted under exactly the
same conditions to verify the reproducibility of the results.
Geometric variations included different inclinations of the
uphill slope from 2:3 to 3:1 (33.78 to 71.68).
The compaction was tested in a first model with 100%
standard Proctor density, in a second with 90% standard
Proctor density and in a third called compaction in
zones, with a looser cushion (90% standard Proctor
density), situated as a buffer in front of a well-compacted
core. All these variations were built up for both reinforced
and non-reinforced tests. Several alternative arrangements
of the geosynthetic elements were investigated by varying
the anchoring length, the barrier type (sandwich or com-
pound), and the facing method.
2.4. Analyses and results
The measured data were analysed, first, by a detailed
comparison of each registered signal (i.e. force and
acceleration during impact) for the first three strokes of
each test, and second by an overall comparison, consider-
ing every stroke until failure of the model.
The detailed comparison made it possible to define a
useful term or characteristic parameter for evaluating the
resistance of earth-fill barriers: Single values, such as the
peak values of force or acceleration, showed specific
dependences on the condition (local compaction) of the
impact area, especially for the first soft impacts of each
test. Therefore the use of a combined term, such as
impulse or energy, seemed more appropriate in this con-
text. Figure 33 gives an example of the forcetime
relation for both reinforced and non-reinforced standard
barriers. Whereas the long impact periods for strokes on a
non-reinforced barrier caused low peak values in the force
signal, the reinforced tests showed the opposite behaviour.
Therefore the area beyond the signal (the impulse) could
be used to reasonably describe the resistance of the
barriers.
The evaluation showed that in this example the trans-
mitted impulse of impacts on reinforced barriers attained
higher values than on non-reinforced ones. This result was
also obtained when comparing other data from reinforce-
ment tests. Analysing compaction tests in detail showed
that proper compaction of the barrier caused better shear
resistance and consequently a higher impulse at impact.
The effects of the absorbed impulse can be clearly
visualised by comparing for example a reinforced model
with its non-reinforced equivalent from the upper camera
position, as shown in Figures 34a and 34b.
The results of the detailed comparison were similar
when evaluating not only the first three strokes but also
the number of strokes that was required to destroy the
model completely. First, the number of strokes differed for
each of the tested variations, and second the measured and
calculated data for each impact yielded different values.
Figure 35 summarises the overall sum of impulses for
each tested model, whereby the value of the non-rein-
forced standard barrier was assumed to be 100%.
Most remarkable was the overall difference between the
reinforced and non-reinforced barriers. Clear relations
with the geometry and the mass could also be obtained:
The steeper the slope inclination, or the lower the mass,
the lower the impulses that were registered.
Lower compaction always caused lower resistance of
the barrier, but the difference from the standard values
was more significant for the non-reinforced models.
Two variations were clearly ahead of all others, namely
the compound type and the sandwich type. Those models
were designed to take the highest loads in two different
ways: The sandwich type required a lot of geosynthetic
reinforcement to cover each layer completely, whereas the
compound type used a relatively small area at the facing,
where the upper and lower ends of the geosynthetics were
bonded.
2.5. Economic efficiency
To include economic matters, cost calculations for the
corresponding prototypes of the model barriers were
carried out. The following assumptions simulated a parti-
cular project
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 353
Geosynthetics International, 2011, 18, No. 6
Test 09
Compaction in zones
9
0
%
4
:
5
10
4
:
5
5
0
135.0
100%
3
:
2
Test 07
Slope inclination 3:1
Test 04
Slope inclination 2:3
Test 06
Slope inclination 2:1
10
2
:
1
4
:
5
5
0
97.5
100%
Test 08
Compaction low
10
10
10
10
10
4
:
5
1
:
1
3
:
1
2
:
3
4
:
5
4
:
5
4
:
5
4
:
5
4
:
5
4
:
5
5
0
5
0
5
0
5
0
5
0
135.0
122.5
89.2
147.5
100%
100%
100%
100%
90%
135.0
Tests 0103
Standard
Test 05
Slope inclination 1:1
Figure 31. Cross-sections of the non-reinforced barrier models; percentage of compaction refers to Proctor standard density;
dimensions of dams in cm
354 Brandl
Geosynthetics International, 2011, 18, No. 6
Test 15
GTX compound type
Test 17
GTX 2:1
Test 12
GTX without facing
Test 14
GTX sandwich-type
Test 16
GTX 1:1
Tests 1011
GTX standard/GGR
Test 13
GTX short anchoring
10
5
10
10
10
10
10
4
:
5
4
:
5
4
:
5
2
:
1
4
:
5
4
:
5
1
:
1
3
:
1
4
:
5
4
:
5
4
:
5
4
:
5
4
:
5
4
:
5
4
:
5
5
0
2
5
5
0
5
0
5
0
5
0
5
0
5
0
135.0
67.5
135.0
97.5
135.0
10
100%
100%
100%
100%
100%
100%
100%
100%
35
8.75
35
35
35
35
35
89.2
135.0
122.5
Test 18
GTX 3:1
Test 19
GTX compaction low
10
4
:
5
4
:
5
5
0
135.0
90%
35
Test 20
GTX compaction in zones
10
4
:
5
4
:
5
5
0
135.0
100%
35
4
:
5
3
:
2
9
0
%
Figure 32. Cross-sections of the reinforced models of barrier fill dams; percentage of compaction refers to Proctor standard
density; dimensions of dams in cm
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 355
Geosynthetics International, 2011, 18, No. 6
barrier length of 200 m, constructed in 0.5 m thick
layers
fictitious construction time of 7 months
0.3 m overlapping of the geosynthetic reinforcement
2 km transport distance for the fill material
no direct passing over the reinforcement by skip
lorries
wedge of humus for a vegetated facing of the
reinforced barrier zone.
For the reinforced structures, obstructions between the
filling and reinforcing working teams on the site were
taken into account: Because of the barriers geometry,
filling with constant efficiency yields an increasing head-
way of the filling crew. To maintain sufficient utilisation
of the expensive filling machinery, the reinforcement crew
was accordingly resized. Even so, a certain loss of
efficiency had to be considered when filling the top
(narrow) layers.
Figure 36 illustrates clear relations between volume and
cost, as well as an approximately 40% higher cost level
for reinforced barriers compared with non-reinforced ones.
Filling without high-quality compaction does not yield
significantly lower cost, because the vibrating roller is
neither an expensive equipment nor on the critical path
regarding efficiency. The remarkably higher level of the
sandwich-type barrier results from the higher amount of
reinforcement in all fill layers.
To outline the economical efficiency of the barrier
variations, a valuecost ratio was finally determined by
dividing the sum of the impulse by the calculated cost.
Figure 37 shows the results:
The generally higher impulse/cost level of the rein-
forced structures emphasises that geosynthetic reinforce-
ment is more effective than additional expense. Bonding
layers for a compound facing provides a much higher
Non-reinforced, 1st impact
Non-reinforced, 2nd impact
Non-reinforced, 3rd impact
Reinforced, 1st impact
Reinforced, 2nd impact
Reinforced, 3rd impact
I
m
p
a
c
t

f
o
r
c
e

(
N
)
Impulse (Ns)
Impact period (s)
900
800
700
600
500
400
300
200
100
0
0 10 20 30 40 50 60 70 80
Impact period (ms)
I
m
p
a
c
t

f
o
r
c
e

(
N
)
Figure 33. Example of forcetime relation for the first three impacts (reinforced/non-reinforced)
(a)
(b)
Figure 34. Cracks in (a) a non-reinforced and (b) a
reinforced model caused by the impact of the rockfall
pendulum
356 Brandl
Geosynthetics International, 2011, 18, No. 6
100
128
80
63
42
66
71
297 298
261
228
348
437
202
162
143
230
262
S
u
m

o
f
i
m
p
u
l
s
e

a
s

p
e
r
c
e
n
t
a
g
e

o
f
s
t
a
n
d
a
r
d

t
e
s
t

(
%
)
Non-reinforced
450
400
350
300
250
200
100
50
0
(
0
1
-
0
3
)

S
t
a
n
d
a
r
d
(
0
4
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

2
:
3
(
0
5
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

1
:
1
(
0
6
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

2
:
1
(
0
7
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

3
:
1
(
0
8
)

C
o
m
p
a
c
t
i
o
n

l
o
w
(
0
9
)

C
o
m
p
a
c
t
i
o
n

i
n

z
o
n
e
s
(
1
0
)

G
T
X

s
t
a
n
d
a
r
d
(
1
1
)

G
G
R

s
t
a
n
d
a
r
d
(
1
2
)

G
T
X

w
i
t
h
o
u
t

f
a
c
i
n
g
(
1
3
)

G
T
X

s
h
o
r
t

a
n
c
h
o
r
i
n
g
(
1
4
)

G
T
X

s
a
n
d
w
i
c
h

t
y
p
e
(
1
5
)

G
T
X

c
o
m
p
o
u
n
d

t
y
p
e
(
1
6
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

1
:
1
(
1
7
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

2
:
1
(
1
8
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

3
:
1
(
1
9
)

G
T
X

c
o
m
p
a
c
t
i
o
n

l
o
w
(
2
0
)

G
T
X

c
o
m
p
a
c
t
i
o
n

i
n

z
o
n
e
s
Reinforced
150
Figure 35. Sum of impulse for all tested models as a percentage of the non-reinforced standard test
C
o
s
t

(
m
i
l
l
i
o
n
s
)

Non-reinforced
16
14
12
10
8
6
4
2
0
(
0
1
-
0
3
)

S
t
a
n
d
a
r
d
(
0
5
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

1
:
1
(
0
6
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

2
:
1
(
0
7
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

3
:
1
(
0
8
)

C
o
m
p
a
c
t
i
o
n

l
o
w
(
0
9
)

C
o
m
p
a
c
t
i
o
n

i
n

z
o
n
e
s
(
1
0
)

G
T
X

s
t
a
n
d
a
r
d
(
1
2
)

G
T
X

w
i
t
h
o
u
t

f
a
c
i
n
g
(
1
3
)

G
T
X

s
h
o
r
t

a
n
c
h
o
r
i
n
g
(
1
4
)

G
T
X

s
a
n
d
w
i
c
h

t
y
p
e
(
1
5
)

G
T
X

c
o
m
p
o
u
n
d

t
y
p
e
(
1
6
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

1
:
1
(
1
7
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

2
:
1
(
1
8
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

3
:
1
(
1
9
)

G
T
X

c
o
m
p
a
c
t
i
o
n

l
o
w
(
2
0
)

G
T
X

c
o
m
p
a
c
t
i
o
n

i
n

z
o
n
e
s
Reinforced
7.5
6.8
5.5
5.1
7.4 7.5
9.4
8.8
8.7
15.1
9.4
8.7
7.4
6.9
9.3 9.4
Figure 36. Calculated cost in millions of euros
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 357
Geosynthetics International, 2011, 18, No. 6
resistance against dynamic impact at comparatively low
cost. The effectiveness of sandwich-type barriers is rel-
atively small, owing to the very high amount of reinforce-
ment; its value ranges even behind the 1:1 inclined
variation with a lower mass. The barriers with low
compaction (tests 08 and 09) rank at last but two and last
but one position. From the detailed comparison and video
analyses it can be concluded that a lack of shear resistance
obviously causes the poor results for these barrier types.
2.6. Practical conclusions from model tests
Twenty qualitative model tests on rockfall protection
barriers provided detailed information about their defor-
mation and failure behaviour. The effects of reinforce-
ment, compaction degree and geometry/mass could be
clearly determined. From the applied measuring equip-
ment, the exact high-dynamic registration of the impact
process provided most significant results, supported by
single-frame analyses of video recordings.
To evaluate the overall resistance of such barriers, the
sum of the impulse proved to be a reasonable term. The
test series on different structures showed a clear advantage
of reinforced barriers, which could also be verified by
comparative studies considering the economical aspects.
In addition to a wider load distribution, reinforced
barriers offer more possibilities for a proper design of the
uphill slope and the protection against overtopping. The
tests emphasised that, predominantly, the layer that was
directly exposed to the impact was stretched. By creating
a compound between the adjacent geosynthetic inclusions
(sewing, welding, bonding, etc.), the upper and lower fill
layers could be additionally activated as load distributors.
If no compound type is used, both the upper and lower
anchoring length should be sufficiently dimensioned,
otherwise heavy dynamic impacts would overstretch the
reinforcement.
Even a simple placement of geosynthetics without cover
on the slope facing (test 12) improves the resistance of the
structure significantly. If a short construction time is
required, this could be a reasonable alternative or compro-
mise.
Comparing just the reinforced alternatives, it has to be
stated that the filling volume/mass cannot be substituted
by a special arrangement of the reinforcement (except for
the compound types). Nevertheless, reinforced barriers of
lower mass showed a higher resistance than non-reinforced
barriers with a higher mass.
Regarding the orientation of the geosynthetic strips,
static and dynamic effects have to be distinguished. The
best orientation from a static point of view is transverse to
the barrier, according to reinforced earth structures or
retaining walls. Overlapping is not then required in the
main direction of tension.
In the case of a rock impact, however, the reinforcement
is strained mainly in the longitudinal direction. A trans-
verse orientation of the geosynthetic strips would therefore
be disadvantageous A compromise is an alternate placing
of transverse and longitudinal reinforcement in subsequent
Non-reinforced
0
(
0
1
-
0
3
)

S
t
a
n
d
a
r
d
(
0
5
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

1
:
1
(
0
6
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

2
:
1
(
0
7
)

S
l
o
p
e

i
n
c
l
i
n
a
t
i
o
n

3
:
1
(
0
8
)

C
o
m
p
a
c
t
i
o
n

l
o
w
(
0
9
)

C
o
m
p
a
c
t
i
o
n

i
n

z
o
n
e
s
(
1
0
)

G
T
X

s
t
a
n
d
a
r
d
(
1
2
)

G
T
X

w
i
t
h
o
u
t

f
a
c
i
n
g
(
1
3
)

G
T
X

s
h
o
r
t

a
n
c
h
o
r
i
n
g
(
1
4
)

G
T
X

s
a
n
d
w
i
c
h

t
y
p
e
(
1
5
)

G
T
X

c
o
m
p
o
u
n
d

t
y
p
e
(
1
6
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

1
:
1
(
1
7
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

2
:
1
(
1
8
)

G
T
X

s
l
o
p
e

i
n
c
l
i
n
.

3
:
1
(
1
9
)

G
T
X

c
o
m
p
a
c
t
i
o
n

l
o
w
(
2
0
)

G
T
X

c
o
m
p
a
c
t
i
o
n

i
n

z
o
n
e
s
Reinforced
100
S
u
m

i
m
p
u
l
s
e
/
c
o
s
t

a
s

p
e
r
c
e
n
t
a
g
e

o
f
s
t
a
n
d
a
r
d

t
e
s
t

(
%
)
350
300
250
200
100
50
150
88
85
61
68
71
237
223
196
172
347
174
164
154
184
209
Figure 37. Value/cost ratio as a percentage of the non-reinforced standard test
358 Brandl
Geosynthetics International, 2011, 18, No. 6
fill layers. Moreover, an isotropic behaviour of the
geosynthetics would be favourable, and various kinds of
connection instead of overlapping should be applied.
Finally, the models tests showed that the zoned dams (e.g.
tests 09 and 20) did not perform better than the homo-
geneous fill dams. A kind of loose cushion zone (buffer)
would possibly cut the peaks of the internal forces, but
decrease the barriers resistance against puncture or break-
through. If barriers are designed with an extremely slender
cross-section to save space, the stability against over-
turning and internal restraining forces of the structure
decrease.
3. GEOSYNTHETICSOIL BARRIERS
AGAINST ROCKFALL, AVALANCHES
AND DEBRIS FLOWS: DESIGN AND
CASE HISTORIES
3.1. Design assumptions
The design of geosyntheticsoil barriers against rockfall,
avalanches and debris flow involves inherently more
uncertainties than for other geotechnical structures, espe-
cially if they are situated in seismic areas (seismic aspects
are not discussed in this paper).
Superimposed on the usual uncertainties regarding
scatter and stress-dependent change of soil parameters and
interactive/composite effects between soil and geosyn-
thetics, is the imponderability of the dynamic impacts,
magnitude and frequency of disastrous events. For in-
stance, rock blocks weighing several tonnes (Figure 27)
with fall heights of 100 m and more have a fall energy
that cannot be absorbed by any protective structure with-
out material plasticization and local damage. Accordingly,
flexible geosynthetic-reinforced barriers or covers have
clear advantages over rigid protective structures. They are
designed after the semi-empirical design method based on
calculated risk and contingency plans (Brandl 1979):
Assuming the worst ground parameters and maximum
credible natural event would make it impossible to achieve
the theoretically required safety factors in many mountai-
nous regions. Consequently, local damage is accepted, but
the possibility to easily repair or strengthen the structure
must be provided. In following this design philosophy, the
safety factors may fall to F 1.0 at the moment of a
rockfall impact causing large deformations. Rock penetra-
tion of 1 to 2.5 m into a geosynthetic-reinforced barrier
dam can be repaired more easily than severely fractured
reinforced concrete barriers.
The so-called acceptable safety factor for permanent
loads must be higher, depending on the risk of failure, the
failure potential and other factors (as in Section 4). The
acceptable safety factor, commonly used for embankment
stability as well as in practically all types of structural
instability, is an empirical method to ensure a level of risk
of failure considered acceptable by the owner, and the
public. After all, no structure is absolutely safe, and the
objective of a design is to ensure an acceptable level of
risk of failure rather than a certain safety factor.
Numerical modelling of rockfalls requires knowledge
about the penetration depth of rock boulders into the
barrier structure, and about the dynamic forces resulting
from the impact. Both depend on the fall height, the
boulder mass, and the indentation resistance of the
structure.
For barrier dams against avalanches, different design
assumptions are relevant than for rockfall protection dams.
They are quite similar to those for barriers against mud-
flows or debris flows. Mass, density and front velocity at
impact are the dominating parameters. Values of more
than 20 m/s have been observed in many cases.
Large-scale tests with avalanches in the field are usually
impossible, owing to the high risk potential. Therefore
computer-aided events are simulated, thus providing valu-
able data for design of barrier systems. There are several
types of avalanche, with widely differing dynamic charac-
teristics: they include high-speed dust avalanches (Figure
38) and lower-speed wet snow avalanches of high density.
Avalanches do not create such excessive local energy
peaks on the barrier facing as huge blocks of rockfall:
there is, instead, a quasi-uniform load distribution. Never-
theless, high puncture resistance of the geosynthetics is
required, because avalanches may include tree trunks,
which can perform like spears.
Common material models in geotechnical engineering
are suitable for describing the behaviour of soils, but not
the complex interaction of various types of soil and
geosynthetics. Therefore reliable numerical simulations
require both adequate material modelling and identifica-
tion of the material parameters involved. The scattering of
material parameters in geotechnical engineering renders
the parameter identification process a challenging task
(Pichler 2003). Because of the non-linear behaviour of the
shear parameters of soil reinforced with geosynthetics, it
is recommended that the parameters be determined under
conditions similar to those in the field (stress level,
compaction degree).
Many soils exhibit a decrease of shear strength with
increasing shear deformation until the residual value (
r
;
Figure 38. Dust avalanche near a high concrete dam
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 359
Geosynthetics International, 2011, 18, No. 6
c
r
0) is reached. This may also occur along soil/
geosynthetic interfaces. Impacts from avalanches and
especially from rockfall are very short. Consequently,
despite potentially large deformations of barrier structures,
the calculation may be based on design shear values of

r
,
design
,
peak
(1)
In most cases a design value close to the peak value is
tolerable, provided that the decrease of the friction angle
and residual shear angle with increasing normal stress is
considered. This refers to soils as well as to geosynthetic
interlayers and interface friction.
Usually, the uphill slope of barrier dams should be
clearly steeper than the downhill one. This provides a
larger catch area, and reduces the risk of crest overrunning
by rock blocks, snow or debris. Additionally, catch fences
are installed on most barrier dams. This causes strong
stress constraints at the crest, requiring particular geosyn-
thetic reinforcement. The fences themselves are occasion-
ally of synthetic material, but usually of high-tensile steel
wire (coated nets and meshes, etc.). At present, the maxi-
mum retention capacity of fences is 5,000 kJ. Synthetic
fences are used only on top of critical mountain zones to
avoid the formation of avalanches (Figure 39).
Figure 40 shows different arrangements for reinforcing
steep soil dams and embankments. Usually, the geosyn-
thetic reinforcement exhibits equal spacing and length of
the geosynthetic inclusions (Figure 40a). An irregular
spacing pattern reflects those cases where stresses are to
be expected higher on the top than in the lower regions
(Figure 40b). Short edge strips (secondary reinforcement)
provide surface protection if the spacing of the geosyn-
thetic layers is large (Figure 40c). A similar effect as in
Figure 40b can be achieved by keeping the layers equally
spaced but varying the length (Figure 40d); short facing
layers serve as surface protection, and also facilitate
surface compaction to achieve high compaction at the
edge of the slope.
The focus in the design of geosynthetic-reinforced
dams, embankments or barrier structures is on internal
and external stability, considering site-specific aspects
such as surface and facing details.
3.2. General stability considerations
There are different possibilities for designing geosyn-
thetic-reinforced barrier dams. In this section a sophisti-
cated method is developed. It was derived from
conventional calculation methods for non-reinforced bar-
rier structures (Brandl and Adam 2000). The design
progresses in steps, as follows.
Internal stability is first addressed to determine the
geosynthetic spacing, length and overlap. Geometry,
surcharge loads, soil parameters such as angle of
internal friction and cohesion, geosynthetic para-
meters, and interaction parameters such as adhesion
between soil and geosynthetic are therefore taken
into account.
External stability calculations against global slope
failure, sliding and base failure (especially on soft
soil) have to be carried out in the next step.
Furthermore, a transition from internal to external
slope stability has to be considered. The usual
geotechnical engineering approach to slope stability
problems is to use limit equilibrium concepts
assuming curved or plane failure surfaces, thereby
yielding an equation for the factor of safety
(Figure 41). The problem can be solved using total
stresses or effective stresses. The use of total stress
analysis is recommended for embankments where
water is not involved, or when the soil is not
saturated. Effective stress analyses are preferred for
conditions where water and saturated soil are
involved. Equation 2 for total stress analysis and
Equation 3 for effective stress analysis describe
limit equilibrium for a geosynthetic-reinforced
dam. Parameters with an overbar represent effective
values, whereas the same expressions without an
overbar are total values. The factor of safety FS
results
FS

n
i1
N
i
tan cl
i
R

m
i1
T
Gi

n
i1
W
i
sin
i
R
(2)
Figure 39. Protective fences on top of a mountain to prevent
the formation of avalanches
(a) (b)
(c) (d)
Figure 40. Various geotextile deployment schemes for
stabilising steep soil dams: (a) evenly spaced, same length;
(b) unevenly spaced, same length; (c) evenly spaced, same
length, with short facing layers; (d) evenly spaced, different
lengths, with short facing layers (after Koerner 1998)
360 Brandl
Geosynthetics International, 2011, 18, No. 6
FS

n
i1
N
i
tan cl
i
_ _
R

m
i1
T
Gi

n
i1
W
i
sin
i
R
(3)
For saturated fine-grained cohesive soils whose shear
strength can be estimated from undrained conditions,
the problem can be simplified. Slices need not be
taken, since the soil does not then depend on the
normal force on the shear plane. Figure 42 shows
details of this situation, which results in equation (4).
FS
cL
arc
R

m
i1
T
Gi

i
W
(4)
Finally, the details of to the surface and facing of the
protective structure have to be considered. When
using geosynthetics with different properties in the
two directions, it is important to recognise how to
place the geosynthetics in an optimum way. For two-
dimensional cases, the maximum stress is typically
in the direction of the dam face. For three
dimensional cases that is, for local impacts
placing in the transverse direction can be more
effective.
3.3. Impact load on geosynthetic-reinforced barrier
dams
The consideration of local dynamic impacts is essential
for protective embankment dams. On the one hand, it is
difficult to predict the area and the magnitude of the
impact force; on the other hand, exact calculation proce-
dures are complicated and costly. In this paper a calcula-
tion method is presented using physical simplifications
and approximations. The procedure is derived from the
simple case of a wedge-shaped dam cross-section, and can
be extended to more complicated cross-section shapes.
3.3.1. Idealisation of fill dam structure
The dam cross-section is idealised as a wedge (Figure 43).
One slice with constant width in the dam axis is consid-
ered. Due to the stocky shape of the vertical beam
bending deformations can be neglected compared with
shear deformations. Accordingly, horizontal load impacts
cause primarily horizontal shear deformations and hor-
izontal shear forces in the dam.
Considering a differential segment Ad of the shear
beam (Figure 44), the differential shear deflection d can
be written as
d
F
A9G
d d (5)
where F is the horizontal shear force; G is the shear
modulus of the dam; and A9 is the corrected area, derived
from dividing the area A by the geometric correction
coefficient k.
Integration of Equation 5, taking into consideration the
boundary conditions as shown in Figure 43, results in
Equation 7 for the horizontal displacement according to
the horizontal force on the top of the dam.

_

F
A9 G
d (6)

Fk
A
A
A
G
ln

A
(7)
The displacements on the top,
0
, and on the foundation
base,
A
, are

0

0

Fk
A
A
A
G
ln

0

A
(8a)

A

A
0 (8b)
The first derivation of Equation 7 corresponds to the
rotation of shear
( , )
n 1
R
Toe of slope
(0, 0)
1
2
3

i
T
Gi
T
G1
T
G2
Slice i
W
i
c
o
s

i
W
i
W
i
i
s
in


i
C
F
N
(b)
(a)
n

Figure 41. Details of circular arc slope stability analysis for


(c, ) shear strength soils (Koerner 1998)
( , )
R
1

i
T
Gi
T
G1
T
G2

W
C.G.
c
Figure 42. Details of circular arc slope stability analysis for
soil strength represented by undrained conditions (Koerner
1998)
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 361
Geosynthetics International, 2011, 18, No. 6
9
Fk
A
A
A
G
1

(9)
The described formulas represent the exact static solu-
tion for the vertical shear beam with variable area
representing a slice of the wedge-shaped dam according
to a horizontal load on the dam crown, which serves as
the fundamental solution for the dynamic behaviour con-
sidered in the following.
3.3.2. RayleighRitz approximation method
The formulation of the dynamic continuum problem
results in a set of differential equations with an infinite
number of degrees of freedom. The basic differential
equations of such distributed parameter systems, with
associated boundary and initial conditions, even in the
actual case of linear elastic solids, but with non-simple
geometry, cannot be solved in an exact manner. In this
case the RayleighRitz approximation method is used to
overcome these difficulties: The essential boundary condi-
tions are implemented in the approximation, which is not
a solution of the basic differential equations. The basic
idea is to approximate the displacement

(, t ) of the
shear-deformable dam by one function separable in space
() and time q(t ), the so-called Ritz approximation
(Ziegler 1998)

, t q t (10)
where q(t ) is the generalised coordinate of the single
degree of freedom (SDOF) equivalent system of the
continuum. The function () is properly selected in order
to meet the requirements of the essential boundary condi-
tions. The function must necessarily comply with the
geometric boundary conditions, and should as far as
possible also take into account any dynamic boundary
condition. In the actual case the function () is selected
from the exact static solution derived in the section above,
which was determined in a sense of best fit for the
dynamic problem.
The function is normalised in such a way that the
deflection on top of the dam is (
0
) 1
ln

0

A
_ _
1
ln

A
(11)
From Equation 11 the normalised rotation of shear is
derived as
9 ln

0

A
_ _
1
1

(12)
The approximated rotation of shear,

(, t ), is then
defined as

, t q t (13)
The original system can be rewritten in an equivalent
system of a Lagrange equation of motion of an SDOF
system, taking energy considerations into account. With
the normalisation of the Ritz approximation, the general-
ised coordinate q(t ) is well illustrated as the measure of
the translational motion of an equivalent mass m

, and the
kinetic energy becomes
E
kin

1
2
_

0
rA _

2
, t d

1
2
m

_ q
2
t
(14)
From Equation 14 the equivalent mass m

can be
determined by using the Ritz separation approximation.
Integration yields
m


rA
A
4
A
ln

0

A
_ _
2
3
2
A

2
0
1 2 ln

0

A
2 ln

0

A
_ _
2
_ _ _ _
(15)
The potential energy is approximated by the strain
Deformation

A

A

0
0

F
0

0
A A
0 0
/
A A

A A ( ) /
A A
d ( )
( )
A
A

A
0
( )
A


A A
( ) 0

0 0
( )
( )
0
Figure 43. Idealised shear-deformable dam cross-section
F A
G
A
d
F
d
Figure 44. Infinitesimal shear-deformable element
362 Brandl
Geosynthetics International, 2011, 18, No. 6
energy of an equivalent spring coefficient k

, deforming
according to Equations 10 and 13, to give
E
pot

1
2
_

0
GA9

2 , t d

1
2
k

q
2
t
(16)
Integration of Equation 16 results in the effective
stiffness k

, given by
k


GA
A
k
A
ln

0

A
_ _
1
(17)
The resulting Lagrange equation of motion of the
idealised dam is that of a linear oscillator with natural
frequency
0
qq t
2
0
q t 0,
0

(18a, b)
Knowledge of the motion behaviour of the dam can be
used for dynamic analyses, that is, earthquake calcula-
tions. In the following, the basic solution will be used to
design a dam loaded by a local dynamic impact, such as a
severe rockfall penetrating into the dam. The dynamic
incident is idealised by an inelastic impact at a point.
3.3.3. Idealised inelastic impact
Impact is a process of sudden exchange between two
colliding bodies within a short time of contact. With
respect to a single impacted body or structure, loading in
such a process acts with high intensity during this short
period of time. As a result, the initial velocity distribution
is rapidly changed. Such rapid loading in the contacting
area is a source whereby waves are emitted that propagate
with finite speeds through the dam body, absorbing the
effective energy.
The most critical case is characterised by a horizontal
impact of a rigid body at the dam crown (Figure 45). The
rigid body, that is, a rock with mass m
S
approaches
the dam with velocity v
S
: A plausible assumption for the
velocity distribution must be made that renders deforma-
tion in the subsequent motion over time. By considering
the static deformation of the linear elastic dam under the
action of a dead weight load F
0
applied at the crown of
the dam, pointing in the same horizontal direction, a
compatible velocity distribution can be assumed that has
the same linear distribution. Consequently, the generalised
velocity after impact can be derived directly from the
RayleighRitz approximation defined in Equation 10 as
_

, t _ q t (19)
In the case of utmost dissipation, it is assumed that the
colliding bodies do not separate immediately after impact.
The surface points of contact take on a common compo-
nent of velocity in the direction of the impact at the end
of the collision process, to give
_ 9


0
, t _ q9 t
0
v9
S
(20a)
or
_ 9
0

_ q9 t v9
S
(20b)
Applying the momentum relation on each partial sys-
tem, taking into account the condition of idealised
inelastic impact (Equation 10), yields the velocity to
common both bodies as
_ 9
0

_ q9 t v9
S

m
S
m
S
m

v
S
(21)
Taking into account the conservation of energy, the
maximum deflection
u
0
q
u
of the dam can be calcu-
lated. Conservation of energy requires
E9
kin
E9
pot
E
u
kin
E
u
pot
(22)
whereby it is obvious that E9
pot
0 at the moment of
collision and E
u
kin
0 at the moment of maximum deflec-
tion, since deformation velocity is zero. Kinetic energy at
the moment of collision is a maximum, and the potential
energy is a maximum at the moment of maximum
deflection
E9
kin

1
2
m

_ q9
2

1
2
m
S
v9
2
S
(23)
E
u
pot

1
2
k

q
u2
(24)
Combining Equations 23 and 24 with Equation 22
finally yields the maximum deflection of the dam accord-
ing to a horizontal impact of a rigid body as

u
0
q
u

m
S
v
S

m
S
m


_ (25)
The maximum deformation function over the total
height of the dam is given by equations 7 and 10
combined with Equation 25
Original structure Equivalent system
v
s
v
s
m
s
m*
k*
q 0
.
Before impact
q t t
u
0 u
( )
Maximum deflection
after impact
V q
s
.
m gV G , ,
Figure 45. Original cross-section of a barrier fill dam and equivalent system loaded by an idealised inelastic single-body impact
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 363
Geosynthetics International, 2011, 18, No. 6

u
q
u

m
S
v
S

m
S
m


_ ln

0

A
_ _
1
ln

A
(26)
3.3.4. Magnification factor
Because of the linear elastic behaviour of the considered
system, it is possible to introduce a magnification factor.
Thus a reasonable design method is provided, since static
calculations can be carried out in a first step, and in a
second step these results can be multiplied by the
magnification factor in order to obtain the maximum
forces (stresses) and deformations (strains) from dynamic
impacts.
It is recommended that static calculations be performed
introducing a horizontal force F
0
on top of the dam, with
magnitude
F
0
m
S
g (27)
Applying this expression to Equation 18a yields the
maximum horizontal static displacement on the top of the
dam as

stat
0

0

m
S
gk
A
A
A
G
ln

0

A
(28)
The magnification factor is defined by dividing the
maximum dynamic deformation (Equation 25) by the
static deformation (Equation 28) to give

u
0

stat
0

v
S

m
S
m


_
A
A
G
gk
A
ln

0

A
_ _
1
(29)
3.3.5. Limit equilibrium design
Following the considerations above, the impact area
should be designed in such a way that the impact load can
be distributed to a larger dam region, so that maximum
forces and deformations can be reduced in the near field
of impact. Nevertheless, non-linear material behaviour and
ultimate stresses and strains should be considered in an
area near the impact zone. Geosynthetic reinforcement
significantly improves the stability against local failure
caused by heavy impacts.
A finite section of the geosynthetic-reinforced dam
according to Figure 46 is considered. The crown of the
dam is horizontally loaded with the static force m
S
g
multiplied by the magnification factor . Equilibrium in
the horizontal direction means that the resulting shear
force is constant in every horizontal plane of the dam. The
constant driving force T
S
is
T
S
T m
S
g

1
k
A G
d
d
const:
(30)
Nevertheless, the shear stress due to the horizontal load
is decreasing with increasing .
In the case of failure, the driving force equals or
exceeds the maximum allowable shear force resulting
from shear resistance of the dam in failure surfaces
(assumed to be vertical) and horizontal surfaces, creating
the body shown in Figure 46. The resistance the of soil
body is composed of two components; the geosynthetics
contribute the third component.
Shear resistance in horizontal plane, T
R
1
(), taking
into account the shear parameters and c
T
R
1

A
0
2
1

0
_ _
rg
0
tan A
0

0
c (31)
The horizontal shear resistance increases with
increasing depth and is a reliable shear resistance
component.
A
0
Z

G
Z

G
l
0
F m
DYN
0 s
.g.

d
T( )
I( )
A( )
I
A
A
A
(a)
Z

G
Z

G
a
i
.tan
a
i
T
2
.1/cos
T
2
.1/cos
T
1

(b)
(c)
Deformation

( , ) ( )
u
0 0 u
u
0
t t q
( ) d /d

u
u
( , ) t t q ( )
u

Figure 46. Limit equilibrium design of a geosynthetic-


reinforced protection dam (the failure body is assumed to be
wedge shaped): (a) cross-section; (b) plan view; (c) maximum
deformation curve
364 Brandl
Geosynthetics International, 2011, 18, No. 6
Lateral shear resistance in vertical planes, T
R
2
(),
taking into account the shear parameters and c
T
R
2

l
A

0
rg tan
1
3

3

3
0
_ _

1
2

0

2

2
0
_ _
_ _

l
A

A
c
2

2

2
0
_ _
(32)
The lateral shear resistance also increases with
increasing depth, but loses effect, depending on the
opening angle due to the total separation of the
failure body from the remaining dam at large
deformations. Therefore the lateral shear resistance
force should not be considered in a limit equilibrium
design. Nevertheless, it serves as a hidden safety
factor covering heterogeneity and local lower shear
parameters.
Geosynthetic reinforcement provides a significant
increase of resistance against shear failure. It serves
as a load distributor: thus the affected soil body
absorbing the impact energy can be assumed to be
significantly larger. Depending on their stressstrain
characteristics, geosynthetic inclusions are mobilised
in different states of impact loading of the dam.
Linear tensile stressstrain behaviour of the geosyn-
thetics is recommended, whereby the resulting secant
stiffness should be in the range of the elastic
modulus of soil, to take into account the strain
compatibility between soil and geosynthetics. In this
case, soil and geosynthetics would be mobilised
simultaneously. Furthermore, different stressstrain
properties and ultimate tensile strength have to be
considered. From Figure 46b it is obvious that the
geosynthetic properties should be equal in both
directions in the case of a local impact. If there are
different properties, the design must be based on the
minimum tensile strength. Consequently, the geosyn-
thetic resistance force, T
R
G
, can be easily calculated
as
T
R
G
a
i
tan min Z
normal
G
, Z
parallel
G

(33)
A limit equilibrium design requires a clear definition of
the safety factor. In the case of a geosynthetic-reinforced
dam it is proposed to use partial factors of safety
S
and

R
: Ultimate strength values can be adapted to allowable
values for the design, as
T
S

S
<
1

R
T
R
1
2T
R
2
_ _
2T
R
G
_
(34)
In Figure 47 the driving forces T
S
due to a dynamic
impact on the dam crest and the resisting forces T
1
R
are
shown in relation to the dam height. In the top area the
safety requirements are not met: thus a geosynthetic
reinforcement must be installed. The design strength of
the geosynthetics can be determined exactly for each dam
region. It is obvious that the geosynthetic reinforcement
must be concentrated in the top region of the dam; both
the length and spacing can be varied, or geosynthetics with
higher tensile strength can be applied.
When a barrier dam is impacted on a locally limited
area, primary shear forces are produced that may possibly
cause failure. Furthermore, in a three-dimensional consid-
eration the dam can also be affected by a global bending
moment due to a local impact, especially in the upper
regions of the barrier. In earthfill and rockfill dams, tensile
and flexural stresses cannot be absorbed, so that failure
may occur due to this kind of load. Geosynthetic inclusions
with high ultimate tensile force placed in the zone of the
dam slope lead to a significant increase in the factor of
safety. Compound the capacity of a high shear force
transfer between geosynthetics and soil is essential.
Geogrids, geotextiles, geonets and geocomposites are
suitable for achieving practicable and reliable solutions.
Furthermore, high friction between geosynthetics and soil
is required in order to transfer the tensile forces into the
soil along the embedded geosynthetics. An approximate
calculation can be performed to estimate the additional
bending moment taken by the geosynthetic layer. There-
fore a finite section of the dam is considered, as shown
in Figure 48. The bending moment can easily be calcu-
lated as
M Z
parallel
G
(35)
whereby () must be estimated; the dam pressure P
DAM
can be approximated by using earth pressure considera-
tions, taking into account the geometry and the overburden
load.
3.4. Design of geosyntheticsoil barrier dams
Geosynthetics reinforcement of a protective dam improves
the resistance of the structure significantly. Slope angles
can be clearly steeper than for non-reinforced barriers.
Furthermore, long-term surface protection is provided, and
the facing can be designed in various alternatives. An
essential advantage is achieved by the load-distributing
effect of the geosynthetics, so that the danger of local
damage and overall failure decreases significantly. The
impacted energy is transferred to a larger dam area, so
that stress and strain peaks are reduced.
Consequently, the design process involves a modifica-
tion to conventional calculation procedures. As shown in
Geosynthetic
reinforcement
required

A
cohesion Coloumbs friction

T
R
G R
/
T F
S
S
DYN
0 S

T T
R
1
S
,
T
R
1 R
( )/
Figure 47. Limit equilibrium design: driving forces T
S
against resisting forces T
R
1
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 365
Geosynthetics International, 2011, 18, No. 6
the previous sections, the design comprises two steps, with
two different kinds of load
continuous dead loads and quasi-static life loads
local heavy dynamic impact loads.
In Figure 49 an example of a geosynthetic-reinforced
protection dam is shown, specially designed to restrain
rockfalls and other impacts. Geosynthetics are applied in
the dam to meet the following requirements
increase of the slope angles (1, 2, 3)
global stability of the dam (1)
distribution of the impacted load (1, 3)
stability against local impacts in the top region (2)
steep slope stabilisation (3)
facing, surface shaping, and surface compaction
assistance (1, 3, 4, 5)
heavy flexural reinforcement covering impacts (4)
light flexural reinforcement covering impacts (5)
reinforcement of weak and/or unstable foundation
soils (6).
The mass distribution of the dam regarding impact
resistance can be improved significantly by installing a
geosynthetic reinforcement: It facilitates the placement of
relatively more mass in the top zone than in the lower
zone of the protective body. Furthermore, the bottom
area of the dam can be reduced to a minimum while the
dam volume is kept constant, serving as an energy
absorber.
If the topography allows various ground plans the most
effective dam shape is that of a convex curvature. Load is
diverted to the abutments by compression in the arch-like
dam body (Figure 50a) and to the dam base, respectively.
In dams with a straight axis, and especially in concave
curved dams, tensile stresses can be caused by horizontal
impacts, which can be taken only by a tensile reinforce-
ment embedded in the outer slope of the dam (Figure
50b).
The height of such protective structures is more or less
Cross-section
Situation

P
DAM
M
M
M( ) ( ) Z
||
G
P
DAM
( )
Z
||
G
Z
||
G
Figure 48. Limit equilibrium design: flexural stresses from
bending moment covered by geosynthetic reinforcement
embedded in the outer zones of the dam slopes. The
maximum reinforcement tensile force depends on the
ultimate dam pressure P
DAM
Impact
Soft foundation soil
6
1
3
2
4
5
6
1
3
2
4
5
Dam stabilisation and impact load distribution
Impact reinforcement
Wall stabilisation, short facing layers, impact load distribution
Slope surface stabilisation and heavy flexural reinforcement
Slope surface stabilisation and light flexural reinforcement
Increase of bearing capacity;
reduction of settlements;
increase of global stability;
prevention of lateral spreading;
reduction of basis deformation
Figure 49. Example of a geosynthetic-reinforced protection dam; effect of different geosynthetic barrier or reinforcement layers
Impact
Geosynthetic
reinforcement
Dam
abutment
(a)
(b)
Load transfer
by friction Tensile and flexural
reinforcement
Impact
Geosynthetic
reinforcement
Figure 50. Different ground plans of geosynthetic-reinforced
fill barriers (protective dams): (a) convex curved dam;
(b) concave curved dam
366 Brandl
Geosynthetics International, 2011, 18, No. 6
unlimited. Depending on the subsoil, conditions heights
up to an order of 100 m (or even more) are possible.
3.5. Case histories
In 1999 a severe rockfall occurred in Tyrol, Austria, that
required the evacuation of about 300 persons because of
expected further rockfalls of about 300 000 m
3
or even
more (1 million m
3
). The protective measures comprised
two reinforced geosynthetic barrier dams, the larger one
having a height of 25 m and a crest length of 170 m. The
fill material was partly gained by excavating an uphill
catch basin (Figure 51), and had to be placed within two
months, always under the risk of further rockfalls. Down-
slope of the barriers the natural slope steepened, thus
requiring detailed slope stability analyses and future
monitoring. Figure 52 shows the top zone of the barrier
dam, which exhibited strong geosynthetic reinforcement,
and also the inserting of a catch fence.
The geosynthetic was a continuous-filament, mechani-
cally bonded (needle-punched), nonwoven geotextile of
polypropylene, strengthened with high-strength polyester
yarns, whereby the orientation of its reinforcement was
bidirectional. The main geosynthetic characteristics are
listed in Table 2.
The compaction of the barrier dams was optimised and
controlled by roller-integrated continuous compaction con-
trol (CCC; see Section 5). Finally, erosion protection of
the barrier slopes was achieved by placing a three-
dimensional mat of extruded polypropylene monofila-
ments, strengthened by an incorporated geogrid.
Figure 53 shows a geosyntheticsoil protective embank-
ment dam against large-scale avalanches in the Austrian
mountains. The upper part of the structure exhibits a
modular (segmental) block scheme consisting of geosyn-
thetic loops. This system acts like a composite body
according to the deadman principle, whereby friction
along the anchor elements is far less important than in the
case of conventionally reinforced soil structures. Conse-
quently, numerous site measurements and observations
have shown that the internal stability of loop-anchored
structures is actually higher than that assessed by conven-
tional calculation. Such walls can be idealised as truss-like
structures, whereby the loops are considered truss ele-
ments under tension, and the soil between the loops and
modular units represents truss elements under compres-
sion. Another calculation method is similar to that of
cofferdams.
The protective dam/structure of Figure 53 was con-
structed in 1981, and has withstood extreme impacts since
then. Long-term monitoring has confirmed its excellent
behaviour.
The worlds largest protective fill barrier against ava-
lanches was recently finished (2010). It is a 600 m long
and 27 m high earth dam of 500 000 m
3
in Tyrol, Austria,
Figure 51. Rockfall barrier (geosynthetic reinforced earth
dam) with upslope catch basin for 300 000 m
3
2
:
3
H 25m
Single reinforcement
Twin reinforcement
7.0
Catch fence
5
.
0
2
.
0
0
.
5
Geocomposites
Detail
8.0
1.0 1.0 0
.
5
Large-scale
rockslide
Erosion
protection mat
4
.
5 6.0 m
Figure 52. Protective embankment dam against large-scale rockfalls and landslides; soil reinforcement on upper part with
geocomposites; slope cover with humus and geosynthetic erosion mat
Table 2. Geosynthetic characteristics to Figures 51 and 52
Property Value
Tensile strength (kN/m) 10
Elongation at break (%) 13
Tensile strength at 5% (kN/m) 30
Long-term design strength (kN/m) 31.3
(FS creep 120 years)
Thickness (mm) 3.0
Mass (g/m
2
) 580
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 367
Geosynthetics International, 2011, 18, No. 6
reinforced with geosynthetics (Figure 54): geogrids of
high-tensile-strength polyester yarns with polymer coating.
Their quality was adapted to the locally prevailing statical
requirements (Table 3)
Lost formwork of galvanised steel mesh was used to
achieve locally a facing of 2:1, and an erosion protection
grid covered the vegetation soil.
The design of high geosynthetic-reinforced earth struc-
tures should consider that friction angle and residual
shear angle
r
decrease with increasing normal stress. On
the other hand, under low normal stress levels, higher in
situ friction angles can be observed than determined in
conventional shear tests. The inclined plane test provides
more realistic results for interface shear strengths, espe-
cially if differentiating between sudden sliding and gradual
sliding (Pitanga et al. 2009).
4. STABILITY OF FLOATING
EMBANKMENTS AND BARRIER DAMS
IN CREEPING SLOPES
4.1. General
Floating embankments for roads and highways, and
(large-scale) toe fills as stabilising counterweights to
unstable or creeping slopes, represent an increasing field
of geosynthetics application. Embankments along creeping
slopes have become a promising alternative to bridges if
Excavation:
catch basin
for avalanches
1
:
1
.
2
5
0.66
6
.
4
0

m
8
.
5
8

m
5.00 m
0.75
1, 2 3
Tensile
straps
1,2 3 4
S1 S2 S3 S4
9.72 m
(a)
III
II
I
Measuring
levels
0.75
O
rig. ground
Slope
to highway
1
:
1
.
5
1
:
1
.
4
3
S5
0
.
5
0
Measuring devices
for tensile forces
Earth Pressure cells
(b)
2
.
5
0
2
.
5
0
2
.
6
5

m
Z1
Z2 Z3
Z4
S1 S2 S3 S4 S5
Figure 53. Protective embankment dam against large-scale avalanches and debris flows; soil reinforcement on upper part with
a loop-anchored wall system: (a) cross-section; (b) plan view with measurement details
Figure 54. Barrier dam against avalanches: 500 000 m
3
:
L 600 m, H 27 m
Table 3. Geosynthetic characteristics to Figure 54
Property Value
Tensile strength (kN/m)
longitudinal 58168
transverse 30
Elongation at break (%)
longitudinal 10.5
transverse 25
Mesh width (mm)
longitudinal 25
transverse 3530
368 Brandl
Geosynthetics International, 2011, 18, No. 6
they are not too high or if they are situated in the toe zone
of an unstable slope (Figures 55 and 56). The deeper the
creeping layer reaches, and the higher the seismic activity
of an area is, the more advantages gained by the use of
geosynthetic-reinforced floating earth structures. Some-
times a combination of flexible and rigid elements may
provide the optimal solution.
A non-reinforced embankment on a creeping slope
would in the long-term fail as a result of stress
constraints caused by locally differential deformations
within ground and fill (internal safety). Reinforced em-
bankments, however, more or less undergo block sliding,
and the internal safety remains widely preserved.
Geosynthetic reinforcement of floating embankments or
other floating fills serves the following objectives
minimising the embankment mass due to steep fill
slopes
creating a composite body that acts like a quasi-
monolith, and can move with the creeping slope
without (relevant) damage
providing sufficient overall stability of the
embankment natural ground system
providing high earthquake resistance.
Monitoring of such structures is therefore inevitable. On
the other hand, they allow construction in areas that
commonly have been considered as too risky. Only the
transition zones from creeping to stable slope sections
require increased maintenance.
Polymer optical fibre (POF) sensors based on the
optimal time domain reflectometry (OTDR) technique
make it possible to measure high strains (up to 40%) fully
distributed along the fibre integrated in geosynthetics for
structures (slopes and retaining structures, dykes, embank-
ments).
4.2. Dominating design parameters
The dominating parameters for designing floating em-
bankments and barrier dams in creeping slopes are shear
strength and groundwater conditions.
Figure 57 shows a histogram that clearly indicates the
important effect of weather on the number and magnitude
of landslides in a particular region: Heavy, long-lasting
rainfalls in spring 1975 caused numerous, disastrous slides
in some regions of Austria, as never experienced during
the past 150 years. They were favoured by a preceding
very wet autumn and heavy snowfalls during winter,
which left the ground soaked like a sponge, and rock
joints already filled with water before the heavy spring
rains began. Above all, Figure 57 raises the question of
worst-case design parameters, and underlines the impor-
tance of semi-empirical design with calculated risk, based
on the observational method and contingency plans. It
would be uneconomic to construct the most expensive
protective structures by assuming throughout and super-
posing the most unfavourable parameters. In many moun-
tainous regions this is even technologically impossible.
Future additional measures even in connection with
remedial works have proved far less costly than a fully
engineered design based on high theoretical factors of
safety.
Risk assessment and stability analyses of slopes should
always involve determination of the residual shear
strength. This is especially important for creeping and
other unstable slopes. Creeping represents a long-term
process with progressively decreasing shear resistances,
unlike the sudden deformations caused by rockfall, debris
flow or avalanches. This deterioration depends not only on
Typically
60
Soil exchange
C
re
e
p
in
g
slo
p
e
Figure 55. Floating embankment in creeping slope
Figure 56. Geosynthetic-reinforced floating embankment,
60 m high, in rugged, steep terrain and seismic zone
N
u
m
b
e
r

o
f
l
a
n
d
s
l
i
d
e
s
250
200
150
100
50
0
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Time (years)
Figure 57. Landslides in Lower Austria between 1953 and
1999. Singular weather conditions in 1975 caused excessive
mass movements
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 369
Geosynthetics International, 2011, 18, No. 6
the soil parameters (mainly grain size distribution, grain
shapes, mineral contents and density), but also on the
degree of saturation, and on the level of effective normal
stress (Figure 58).
Consequently, if the normal stress in shear tests is too
small, the measured value of
r
is not the theoretical
minimum. As
r
of soils and geosynthetics mostly de-
creases with increasing normal stress, the overburden
should be taken into account when assessing the possible
residual shear strength in the field. Deep-seated slide
planes are more critical than those near to the surface.
An increasing degree of water saturation favours the
tendency towards slickensides and decreasing
r
(Figure
58). Therefore shear or triaxial tests should be performed
on saturated specimens to obtain the minimum value of
r
for lower border analyses.
The rate of displacement has an influence on the
residual strength of a range of soil types, fillings and
weathered decomposed rock with a high proportion of
fines. In granular soils or rock fills, the effect of rate of
shearing on the ultimate strength is negligible. In cohesive
material different behaviour may occur when a shear zone,
formed at a residual strength by slow drained shearing, is
then subjected to more rapid rates of displacement.
Creeping soil slopes close to the limit equilibrium
(F 1), and exhibiting a low residual shear strength, tend
towards progressive failure with a gradual transition from
creeping to (sudden) slip failure. The risk of tertiary creep
increases with decreasing
r
: Long-term monitoring is
therefore essential for a reliable risk assessment, and to
start stabilising, strengthening or retaining measures in
time. If sufficient data exist, a creeping factor can be
deduced, and future extrapolation is possible (Figure 59).
In the case of geosynthetic-reinforced structures in
creeping slopes, or at the toe of unstable slopes, slip
surfaces running through the ground and the structure
have to be investigated. Therefore the interface shear
strength (peak and residual value) between geosynthetics
and fill material, and the composite shear strength of the
reinforced fills (compound body shear strength), should be
determined. Floating embankments and barrier dams in
creeping slopes may undergo relatively large deforma-
tions. Consequently, residual strength is an essential para-
meter for risk assessment and proper design.
4.3. Creeping pressure on barrier dams in unstable
slopes
The main factors influencing the creep rate of an unstable
slope are
slope angle
pattern of discontinuities, rock joint fillings
shear parameters
water pressure
external loads or unloading.
In a slope undergoing creep, retaining structures may be
stressed by a lateral pressure E
cr
, which significantly
exceeds the theoretical earth pressure at rest, E
0
: This
creep pressure, E
cr
, may also be considered as sliding
pressure or stagnation pressure on a retaining structure.
If we assume a (quasi) cohesionless mass and limit
equilibrium , the creep pressure becomes a special
case of an increased Rankine earth pressure according to
(Figure 60)
E
cr
m
1
2
h
2
cos (36)
where is the slope angle; h is the height; and is a
fictitious friction angle, including the effects of water
pressure (and small cohesion; Brandl 1980).
The multiplication factor m() also depends on the
stiffness of the retaining structure. Hence geosynthetic soil
structures attract less creeping pressure than concrete
walls. The enveloping values of Figure 61 were obtained
from numerous in situ measurements. They have already
proved very useful for practical design in sliding areas of
Austria and elsewhere for more than 35 years.
25
20
15
10
5
0
R
e
s
i
d
u
a
l

s
h
e
a
r

a
n
g
l
e
,
(
d
e
g
r
e
e
s
)

r
0 200 400 600 800
Normal stress, (kN/m )
n
2
Degree of
saturation,
(%) S
r
20
40
60
80
95
Figure 58. Residual shear angle
r
against effective normal
stress
n
; degree of saturation S
r
as parameter. Results of
direct shear tests with silty-clayey soil
5 10 50 100 500 1000
1
2 3 5 10
20
5
10
15
20
25
30
35
S
l
o
p
e

m
o
v
e
m
e
n
t
,
(
c
m
)

x
log time, (weeks) t
Slope 1
11.8 cm k
cr

Slope 2
37.4 cm k
cr

(years)
k
cr

x x
2 1

log / t t
2 1
1
Figure 59. Steady creeping behaviour of two unstable slopes
in weathered schists; definition of creeping factor
370 Brandl
Geosynthetics International, 2011, 18, No. 6
4.4. Stability analyses of geosynthetic-reinforced
earth structures in slopes
Stability analyses of geosynthetic-reinforced earth struc-
tures in slopes are based mainly on slope failure calcula-
tions, taking into account the retaining forces in the
reinforcing layers and inclusions. Frequently, external/
global and internal stability are considered entirely sepa-
rately, in a rather formal way: Global failure mechanisms
commonly assume slip surfaces that run completely out-
side the reinforced earth structure, and thus do not cut the
geosynthetic layers or inclusions. The analysis of internal
safety, however, is frequently based on failure lines
running only within the reinforced earth or slope.
In daily design practice, such a formal separation of
stability analyses frequently results in investigations being
limited to a very narrow scope, namely failure mechan-
isms just outside or (predominantly) inside the reinforced
part (e.g. a combination of two sliding bodies with plane
slip surfaces). In many cases, an investigation of potential
failure mechanisms of various shapes and positions run-
ning partly outside and partly inside the reinforced zone is
not carried out.
However, it is specifically such a failure mechanism
often called mixed mode or compound mode that
frequently turns out to be the most probable form of
failure, thus providing the lowest factor of safety with
regard to slope stability. Ignoring this may lead to a
critical underdesign of the structure, and hence to an
increased risk of failure.
Therefore the length of the reinforcement should be
calculated from earth pressure theory as well as from
slope stability analyses. A proper design has to consider
all possible slip surfaces to gain the most critical failure
mechanism. This means that cylindrical slip surfaces have
to be investigated as well as plane, logarithmic, or any
form of combined or polygonal failure surfaces, including
potential slip surfaces in the ground or within the
geosynthetic-reinforced earth structure (e.g. pre-existing
or fossil slip surfaces in the ground, geosynthetic/soil
interfaces, and structural interfaces). An exclusive reliance
on conventional stability analyses, proving internal and
external stability without considering mixed forms of
failure (compound modes), is totally insufficient, as
numerous failure histories have shown.
4.5. Embankments instead of bridges in creeping
slopes
Modern compaction equipment, optimisation and control,
and geosynthetic reinforcement have opened the possibi-
lity of constructing high embankments instead of bridges
for roads, highways and railways (Figure 62).
Experience has shown that generally valid criteria for
weighing up the advantages of embankments in compari-
son with bridges are rather limited. Whether a bridge or
an embankment is most suitable has to be specifically
decided for each project. The main factors that influence
the pros and cons of an embankment instead of a bridge
are
local situation, including existing buildings or
settlements
geomorphology; stability of existing slopes
ground properties, including ground and slope water
conditions
depth of slip surface(s)
seismic activity
allowable total and differential settlements of the
Profile
of creep velocity
N
V
A
Creeping
slope
D

1
M
V

M
B
H
B
B V
B
E
cr
E
cr

c
G
A
D
Q
Slide
plane

Q
G
V
Retaining
structure
h
Figure 60. Theoretical assumptions for calculating the
creeping pressure E
cr
on retaining structures or protective
embankment dams (barrier fills) in a creeping slope. V
B
, H
B
,
M
B
are external forces on the top of the structure
4
3
2
1
M
u
l
t
i
p
l
i
c
a
t
i
o
n

f
a
c
t
o
r
,
(
)
m

E m
cr,h
( ) cos
2
for
20 30 40
Friction angle, (degrees)
h
2
2
R
i
g
i
d
F
le
x
ib
le
h height (depth)
of creeping mass
Figure 61. Influence of stiffness of structure on the creeping
pressure E
cr
; slope angle, (fictitious) friction angle
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 371
Geosynthetics International, 2011, 18, No. 6
embankment crest (with regard to the evenness of
road pavements or rail tracks)
availability of proper fill material
material balance of soil excavation or slope cut and
of fill volume within a certain construction section
length and quality of access ways for the transport of
fill materials
number, diameter, length and location of possible
culverts in the bottom of the embankment or within
the fill
statical system of the bridge
schedule of construction operation
construction costs
costs for long-term maintenance
local climate; environmental and aesthetic aspects.
In the case of statically very sensitive slope bridges
(e.g. with continuous girder superstructures), the founda-
tion requires a high resisting moment (e.g. large-diameter
sockets or caissons), sometimes with multiple permanent
anchorage. This means that rather rigid (and deep)
footings have to be designed.
Moreover, such buildings must be protected upslope by
a flexible retaining structure, which acts as a first barrier
(primary retaining system) against excessive slope pres-
sures (e. g. Figure 63). This may require very long anchors
(up to 100 m or even more) for fixing them in stable
ground. As slope pressures may change with time, long-
term monitoring of sensitive structures in creeping slopes
is essential.
To sum up, bridges along creeping slopes are very
expensive, have a relatively high risk potential, and require
continuous monitoring and comprehensive maintenance.
Moreover, they are more sensitive to earthquakes than
geosynthetic-reinforced embankments.
When comparing bridges and (reinforced) embankments
along creeping slopes, the following aspects should be
considered.
Embankments may replace bridges if the fill does
not significantly increase the slopes creep rate, and
the expected deformation remains within an allow-
able limit value.
The construction costs of (geosynthetic-reinforced)
embankments are smaller than those of bridges.
Commonly, conventional embankments are superior
to bridges only in connection with toe-weighting of
the creeping slope: that is, if the local geomorphol-
ogy allows a toe fill as counterweight. However, in
the case of very deep-reaching creeping strata,
floating earth structures are superior, because bridge
construction and monitoring become uneconomical.
In canyon-like rugged terrain, embankments have
proved suitable if they are intensively teethed with
the original ground. Geosynthetic reinforcement is
recommended then, at least locally (e.g. Figures 56
and 70).
Using lightweight products as fill material reduces
the load applied on creeping slopes, but such
materials exhibit low earthquake resistance, because
they tend to liquefaction (and embankment spread-
ing), especially if they have a uniform grain size
distribution. Moreover, the composite effect between
fill and geosynthetic reinforcement decreases with
lower density.
Geosynthetic reinforcement makes a reduction of the
fill mass possible, owing to the steeper embankment
slopes. Furthermore, such composite earth structures
have a significantly higher resistance against earth-
Slope bridge (conventional design)
Highway
Deposit
Original surface
Embankment (green design)
Village
Figure 62. High embankment instead of a bridge for a
highway or railway. Scheme illustrating the acoustic and
visual screening of a nearby village through proper
vegetation of the embankment slope. Also indicated is the
possibility of a steeper embankment slope due to geosynthetic
reinforcement and of a deposit upslope of the embankment
Original surface
Anchored
wall
Slope wash
Beam
25.5 m
7
0
.
0

m
1
3
.
0

m
7.0 m
1
7
.
0

m
Mica schist
decomposed,
mylonitic zones
E
xte
n
so
m
e
te
rs
1
2
a
n
ch
o
rs
Figure 63. Typical foundation and permanent anchoring of a
bridge pier in a steeply inclined, creeping slope. Statically
sensitive superstructure (2.6 km long bridge) completely
separated from the slide-protection measures (tied-back
retaining structures) around each bridge pier
372 Brandl
Geosynthetics International, 2011, 18, No. 6
quake, spreading and local failure. The overall slope
stability is higher than for conventional fills, but
nevertheless the creep rate of an unstable slope will
increase unless drainage measures are successful.
Dowelling of the creeping zone with large-diameter
piles or reinforced concrete sockets beneath the
embankment is another alternative to slope bridges
(Figure 64).
Running along creeping slopes or crossing them
requires contingency plans for bridges as well as for
embankments. The structures should exhibit the
possibility of future strengthening or relevelling if
the results of long-term monitoring require it.
The maintenance requirements of embankments is
clearly lower than for bridges.
Embankments are environmentally more friendly
than bridges, and their slopes can be planted. In
many cases the earth structure finally looks like
natural terrain (e.g. Figures 65a, 65b and 66).
4.6. Long-term performance of high embankments
or barrier dams in creeping slopes
More than 35 years of personal experience with numerous
embankments and barrier dams of 30 to 135 m height and
a length of about 150 to 700 m can be summarised as
follows.
Usually, the stability factors against slope failure or
ground failure of high embankments are at their lowest
just at the end of construction. Excessive pore water
pressures may occur in fine-grained, wet-fill zones or in
the natural ground. In the long term a gradual decrease in
safety is possible if
Highway
Granular
surface drain
2
:
3
Soil exchange
Drainage
trench
Potential
slide surfaces
Nonwoven
geotextiles
Sliding loam
E
v. a
n
ch
o
rs
Clayey
slickensides

r
5
D
e
p
t
h

t
o

4
5

m
Road
Retaining
panels
5
.
5

m
6.0 m
6.4 m
Rip rap
(cemented)
Sockets (caissons)
6.0 4.5 m
Shotcrete
20 cm d

4
6

m
O
rig
in
a
l g
ro
u
n
d
Figure 64. Highway in an unstable slope with slope dowelling and geosynthetic-reinforced embankment. Cross-section through a
socket wall that exhibits reinforced concrete panels on top of the sockets (caissons) as contingency plan for the possibility of
later tying back with prestressed anchors
(a)
(b)
Figure 65. Highway embankment, 100 m high, instead of a
slope bridge (with comprehensive retaining measures) on toe
of an unstable slope: (a) in 1980 (immediately after
construction); (b) in 2010
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 373
Geosynthetics International, 2011, 18, No. 6
the drainage systems fail
low-quality fill material was used, and insufficiently
compacted
the fill material or the natural ground tend to
progressive failure, owing to a very low residual
shear strength.
Slope stability assessment of high embankments should
consider a possible decrease of the friction angle in the
lower fill zone, depending on the grain size distribution,
grain shape and strength of the fill material: High over-
burden causes grain crushing, thus leading to a flattening
of the MohrCoulomb rupture line in the diagram.
Interface friction between geosynthetics and soil usually
also drops with
n
:
High embankments should be designed with benches at
about 20 to 50 m of vertical spacing, depending on the fill
material, slope angle, slope stability and local climate.
These benches should have a width of at least 3 m. They
serve for maintenance, facilitate local strengthening (if
necessary), and act as a braking zone in the case of local
slope sliding or soil liquefaction (e.g. during heavy rain-
falls or earthquakes).
Benches as well as the toe zone of slopes exhibit local
stress concentration (Figure 67). Therefore these parts
should be reinforced with geosynthetics (Figure 68). In
seismic zones the installation of geosynthetic-reinforced
zones parallel to the sloped surface of high embankments
is recommended, with width at least 10 m. Barrier dams
frequently should be reinforced over the entire area.
Moreover, it can be useful to install reinforced inter-
layers covering the entire ground plan of an embankment,
for example for slender embankments along steep slopes
and without a toe buttress towards the opposite site of a
valley, or without the possibility of a counterweight fill in
a valley bottom.
Such reinforced interlayers should be at least 3 m thick
and be placed at a vertical spacing of about a quarter to a
third of the maximum embankment height H, and on top
of the embankment. They serve as load-distributing
members (reducing differential settlements), and increase
slope stability.
In the case of proper fill material and careful compac-
tion, the self-settlements (ss) of high embankments (i.e.
without ground settlements) typically lie below 1% of the
embankment height H. If high-quality fill material is
available and intensively compacted, self-settlements may
even be reduced to about ss 0.10.2% of H. On the
other hand, poor compaction and too many silty-clayey
interlayers (e.g. placed in a sandwich-like mode) may lead
to self-settlements of about ss 12% of H or more,
depending on the proportion, thickness and quality of
soft interlayers.
Spreading of high embankments is negligible if they are
properly interlocked or toothed with the natural ground
(by cutting steps).
Long-term experience with numerous highway embank-
ments has shown that such earth structures often exhibit
Figure 66. Embankment, 120 m high (large-scale
counterweight), instead of a highway bridge in a creeping
slope, 30 years after construction
Bench or slope toe
Figure 67. Stress concentration at the bench or toe zone of a
slope: results from photoelastic model tests (Hoek and
Gralewska 1969)
Geosynthetic reinforcement
embankment or barrier dam
Erosion
protection
mats
q
Bench zone
reinforcement meets
stress constraints
Figure 68. Recommended geosynthetic reinforcement of
bench zones of high embankments or barrier dams in
unstable slopes; schematic
374 Brandl
Geosynthetics International, 2011, 18, No. 6
significant advantages over bridges. They are environmen-
tally friendly, they facilitate a balance of cut and fill
volume during construction, and they require only negli-
gible long-term maintenance. If proper fill material is used
and carefully compacted, the differential settlements of the
embankment crest, and hence also of the road pavement,
are usually smaller than for low embankments on soft or
heterogeneous ground.
Until about the early 1990s, continuous compaction
control (CCC) and roller-integrated compaction optimisa-
tion were still in an early development phase, and not
generally available. Nowadays, improved compaction
equipment, technology and control methods clearly facil-
itate the construction of high embankments for highways
(and even for high-speed railways). This could be a cost-
effective, environmentally friendly alternative to bridges
that, at the same time, also provides increased factors of
safety.
High embankments crossing unstable slopes in seismic
areas should have local reinforcement with geosynthetics.
Such inclusions have proved very suitable, as they sig-
nificantly improve the internal and external (overall)
stability of the embankments. Moreover, they reduce
differential settlements of the embankment crest.
4.7. Case histories
Figure 69 shows the peripheral zone of a 120 m high
embankment that was constructed instead of a slope
bridge along an unstable slope. The low slope stability
required local reinforcement with geosynthetics; also,
nonwoven geotextiles were placed in the base to provide
sufficient long-term function of the drainage blanket. This
earth structure was constructed between 1978 and 1979;
now it can hardly be distinguished from the natural slope
(see Figure 66). Maintenance is minimal (unlike to the
adjacent slope bridges), and the stability of the natural
slope could be increased significantly.
Floating embankments instead of bridges have proved
suitable also in rugged, steep terrain. Figure 56 gives a
partial view of a 60 m high reinforced embankment in a
seismic zone. The curved crest connects twin tunnels on
either side of the fill, which is intensively toothed with
the natural ground. Earthquake-induced deformations
would therefore be of a more minor effect than in the
case of a bridge. Figure 70 illustrates the situation at the
end of construction. To gain sufficient internal and
external stability, heavy metallic reinforcement (tensile
strength 400 kN/m) was also placed locally.
Geosynthetic reinforcement of floating embankments or
barrier dams becomes especially important in areas of
high seismic activity, and if the shear strength of a
creeping slope is already progressively decreasing towards
the residual value. This is demonstrated by a spectacular
case history.
A 100 m high slope cut for a new highway triggered a
landslide, superimposing old creeping. The moving area
finally had a length of about 600 m, a width of 270 m, a
height difference of 330 m, and a total mass volume of
13 3 10
6
m
3
: Inclinometers disclosed the dominant slip
surfaces at a depth of 83 m, and displacements of 20 mm
per month there. Furthermore, multiple slip surfaces had
to be considered. Geological observations identified fis-
H
1
2
0

m

Original
ground
Weathered
schists
Possible slip surfaces
Benched
Decomposed
wet colluvium
Rockfill
(riprap)
Reinforcement
30 m
Highway
2
:
3
1
:
2
Rockfill
Extremely heterogeneous fill
material (sandy silt to rockfill)
placed in sandwich form (not mixed)
Figure 69. Side zone of a 120 m high embankment instead of a highway bridge. Earth structure serves simultaneously as
counterweight for the creeping slope and is intensively toothed with the natural ground
Figure 70. Geosynthetic-reinforced floating embankment,
60 m high, for a highway interchange in rugged, steep
terrain. Gabion facing in the upper zone using sandstone
from the adjacent tunnel excavation. Geosynthetics also used
to stabilise the cuts in the natural slope (see background)
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 375
Geosynthetics International, 2011, 18, No. 6
sures and open cracks up to 2 m wide in the order of
300 m upslope of the crest of the slope cut.
The entire mountainside where this bit cut was
excavated comprises a disordered, chaotic mass in which
very weak ophiolites are erratically distributed. Moreover,
the ophiolites are serpentinised, sheared, weathered, and
with passages transformed to clayey material. A certain
bedding seems to be dipping into the slope at a gentle
angle of approximately 108, but it is intensively tectonised,
and severely cut by step-like discontinuities (Figure 71).
Numerous discontinuities, more or less parallel to the
slope or cut surface, respectively, govern the local and
overall stability. Moreover, clayey interlayers (mylonites)
with low residual shear strength contributed to the in-
stability of the big cut.
Several alternative alignments were discussed for safe
crossing of the geologically sensitive and seismically
active area of this big cut. Owing to the uncertainties of
a deep-seated landslide the existing alignment and tunnel
options were excluded.
Given the magnitude of the instability and earthquake
danger, the proposed solution in this area was to shift both
carriageways away from the toe of the big cut as far as
possible, taking into account the highway geometrical
restrictions deriving from the presence of existing tunnels
constructed on both sides of the landslide. This required
comprehensive measures to reach acceptable safety fac-
tors, whereby geosynthetics were used for reinforcement,
drainage/filters and erosion protection (Figure 71)
local slope dowelling (shear keys) with large-
diameter bored piles, barrettes or elliptical sockets
on the toe of the big cut
backfill of the toe zone of the cut, and an additional
counterweight fill of 500 000 m
3
, with each layer
over the ground view fully reinforced with geosyn-
thetics, to restrict mainly superficial slides but also
deep-seated slides
partial backfill of the slope cut uphill of the
counterweight to avoid local slides (the toe zone near
the berm should be fully geosynthetic-reinforced)
a counterweight embankment (the big embankment
in Figure 71) of about 3 3 10
6
m
3
, 700 m long and
maximum 135 m high to increase the overall stability
of the creep area, to prevent deep-reaching slips, and
to carry the highway after realignment
an arch-shaped geosynthetic-reinforced support dam
on the toe of the big embankment where the slope
forms a canyon-like valley
comprehensive drainage measures (deep-drainage
borings, drainage trenches, drainage blankets), using
filter geotextiles
open cracks treatment, slope reshaping
contingency plans to respond immediately to critical
monitoring results.
The counterweight fill, upslope backfill and downslope
big embankment were placed on granular drainage blan-
kets, using filter geotextiles if locally necessary. Unlike
Ophiolites
serpentinised, sheared,
weathered, clayey
(bedding not relevant)
Drainage
boreholes
(filtered)
Backfill
Shear key
structure
(old pile rows
integrated)
Counterweight fill
500000 m
(reinforced with geosynthetics)
3
Big embankment
3 10 m
135 m

6 3
max
H
Old highway axis
New highway axis
C.Co
C2
n
C2
C2
n
n
n
n
C2
C2
C2
n
n
C1
n
C2
C2
?
n
?
n
O.M
Se
Se
Se
Se
Se
Se
Se
Se
?
?
O.M
O.M
O.M
O.M
Fk?
n
C2
TE
TE
TE
Figure 71. Geotechnical cross-section through a landslide triggered by a 100 m high slope cut (big cut) and stabilising
measures: counterweight fill and upslope backfill fully reinforced with geosynthetics, big embankment includes only local
geosynthetics
376 Brandl
Geosynthetics International, 2011, 18, No. 6
the upslope fills and the downslope arch dam supporting
the big embankment, the latter was reinforced only
locally, in the bench zones, in sandwich fashion at
vertical-spacings of about a quarter of the maximum
embankment height H, and within the top 5 m of the
embankment. The upper reinforcement served mainly for
stress distribution and differential settlement reduction.
The structural reinforcement consisted of flexible, high-
strength geogrids manufactured from high-tenacity polye-
ster yarns with low creep and an environmentally inert
coating. The longitudinal tensile strength of the upper
reinforcement and the structural reinforcement is 45 and
110 kN/m respectively, and the elongation at break is
12.5%. Additionally, three-dimensional reinforcement
grids with erosion protection were placed on the slope
surface.
Roller-integrated continuous compaction optimisation
and control (CCC) was recommended for all earthworks.
5. OPTIMISED COMPACTION OF
GEOSYNTHETICSOIL SYSTEMS
The interaction between geosynthetics and soil (or other
granular fill material) depends mainly on the interface
shear strength, the stiffness ratio, the geometry of inclu-
sions, and the compaction of the fill layers. A high and
uniform degree of compaction favours composite behav-
iour, thus improving the bearing-deformation character-
istics of the entire system. In addition, homogeneous
compaction avoids local stress concentration and stress
constraints in multi-layered composite systems consisting
of geosynthetics and soil or other granular material.
A high and uniform compaction is especially important
for reinforced soil barrier dams, floating embankments,
counterweight earth structures and other retaining struc-
tures, but new dykes and dams or their enlargement should
also be compacted intensively and uniformly.
So far, compaction control has been carried out mainly
by means of spot test methods (on selected points) to
check the density or stiffness of the compacted layer.
Conventional tests to determine density, and load plate
tests for checking soil stiffness, are based on spot check-
ing by more or less random selection, and they are only
superficial. The uniformity cannot be approved, and weak
points are unlikely to be detected by such spot tests. This
involves an unavoidable residual risk. Moreover, all spot-
test methods are relatively expensive and time consuming.
Finally, conventional testing frequently delays construction
work, because construction activities must not be carried
out in the vicinity of a spot test, as ground vibrations
might affect the test results.
Therefore the conventional methods of compaction
control are no longer sufficient for high-quality projects,
especially for high embankments along steep or unstable
slopes and in seismic areas, for barrier fills, and for dams.
Increasing demands on engineered earth structures require
as much continuous compaction optimisation and control
as possible during the compaction procedure. The roller-
integrated CCC-technique, which was first used for road
structures (Brandl and Adam 1997), represents a distinc-
tive improvement, because all control data are available
during the compaction process and over the roller-com-
pacted area. Moreover, CCC allows a reduction of the
number of conventional acceptance tests, and provides all
data during compaction. Thus the compaction procedure
can be optimised and expedited.
Composite structures of geosynthetic-reinforced soil
require intensive and homogeneous compaction, whereby
the subgrade should be compacted properly (as far as
possible). Theoretical investigations, field experiments
and site measurements have shown that there is an
intensive interaction between the soil or other granular
material (fill layers), the geosynthetic inclusions, and the
compaction equipment. Measuring this interaction pro-
vides an excellent tool for three important goals of
compaction.
Compaction optimisation. This refers to the quality
of compaction, to the required compaction energy
and time, and to the required geotechnical para-
meters of the compacted material. Over-compaction
and re-loosening of layers should be avoided just as
much as heterogeneous degrees of compaction.
Therefore a main goal of the cooperation between
geotechnical and mechanical engineering has been
the development of intelligent compaction equip-
ment that reacts to locally varying soil or granular
material properties by automatically changing its
relevant machine parameters. Rollers with automati-
cally regulating compaction systems (vibratory or
oscillatory) are a significant step in this direction,
which raises compaction from a mere routine craft to
a scientifically based high-technology process.
Compaction documentation. The CCC technology
involves an automatic registration of all data in such
a way that the results cannot be manipulated. This
data collection is essential not only for site
acceptance but also for quality control and long-term
risk assessment. Furthermore, such information is
very helpful for future rehabilitation of dams and
embankments (after floods, earthquakes, etc.).
Compaction control. This should be widely per-
formed during the compaction procedure. A calibra-
tion of the control data based on the reaction
between ground and compaction equipment is
essential. Control tests after compaction should be
increasingly reduced to conventional spot checking,
whereas continuous compaction control (compaction
equipment-integrated) should be promoted.
Soil properties inherently vary within a fairly wide
range. Furthermore, locally different fill materials and
compaction degrees on the construction site, together with
test uncertainties, increase the scatter of compacted fill
material characteristics. Consequently, reliable quality
assessment and stability analyses may become quite
difficult. Such a situation can be significantly improved
by applying roller-integrated continuous compaction con-
trol (CCC), which facilitates compaction optimisation
during the compaction procedure:
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 377
Geosynthetics International, 2011, 18, No. 6
Vibratory roller compaction takes place by means of a
vibrating drum, which is excited by a rotating mass.
Oscillation of the roller drum changes, depending on the
soil response. This fact is used by CCC in order to
determine the stiffness of the ground. Accordingly, the
drum of the vibratory roller is used as a measuring tool
(Figure 72). Its motion behaviour is recorded by the
sensor; analysed in the processor unit, where a dynamic
compaction value is calculated; and visualised on a dial or
on a display unit, where data can also be stored. Further-
more, an auxiliary sensor is necessary to determine the
location of the roller. By means of GPS (global position-
ing system) the position of the roller can be located up to
an accuracy of 5 cm.
The dynamic compaction values have to be calibrated
on the basis of conventional tests: for example, compac-
tion degree D
Pr
, density r, or deformation modulus E
v
(E
v1
rather than E
v2
values from static load plate tests or
E
vd
from dynamic load plate tests). The main advantages
of this control method are
continuous control of the entire area
results available during the compaction process, with
no hindrance or delay of the construction work
optimisation of the compaction work, including
prevention of local overcompaction (which causes
near-surface re-loosening of the layer)
full and permanent documentation of the entire area.
CCC possesses the essential advantage that the measur-
ing equipment can be easily mounted on vibratory or
oscillatory rollers (smooth rollers or sheepsfoot rollers).
Experience has shown that the roller operators and site
supervisors have very quickly familiarised themselves with
this control method. Low-quality rollers, which provide
only low compaction quality, can be eliminated, and the
documented data cannot be manipulated. CCC has proved
suitable on many construction sites, and has therefore
been obligatory for several years already in Austria,
mainly for roads and railways, embankments, dykes and
dams, barrier fills, geosyntheticsoil structures, and clay
liners of waste deposits.
Furthermore, the measuring depth of roller-integrated
CCC is significantly larger than that of conventional
methods: Whereas density measurements commonly reach
a depth of only 0.1 to 0.3 m, and standard load plate tests
about 0.5 to 0.6 m, CCC reaches to a depth of about 2 to
2.5 m.
Figure 73 illustrates this with a case history: The raft
foundation of a power plant required deep soil improve-
ment and partial soil exchange to minimise differential
settlements. Soil improvement of the loose river sediments
(sandy silts) was performed with stone columns, which
were then covered by a geotextile and layers of sandy
gravel. The exact location and diameter of the stone
columns was registered during the entire earthwork, and
could still be observed by CCC on the top layer, which
covered the stone columns by 1.5 m (Figure 73).
Consequently, CCC could easily detect those areas
where geosynthetic reinforcement had locally been
omitted in the sublayer(s) despite the design requirements.
High-quality compaction provides numerous advantages
for dykes, flood protection and barrier dams, floating
embankments, counterweight fills and retaining structures
with and without geosynthetic reinforcement. Moreover, it
has proved successful for roads, highways, railways and
all kinds of earthworks and reinforced retaining structures.
Compared with conventional spot checking CCC provides
increased safety
improved serviceability
improved seismic response
increased lifetime
reduced costs and maintenance
reduced construction time
detailed monitoring of the top zone (2.5 m) of
existing dykes or dams (see Figure 10).
Furthermore, CCC increases the installation survivabil-
ity of geosynthetics placed in multi-layered structures. The
uniformly compacted, smooth surface of each fill shift
reduces the risk of wrinkles or punch-through in the
Display unit
Drum
Sensor
Off-centre
rotating
mass
Processor
Auxiliary sensor
Figure 72. Principle of roller-integrated continuous
compaction control (CCC) and compaction optimisation
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035
1000
800
600
400
200
0
Vibroflotation points
O
m
e
g
a
Limit for sufficient
compaction
Excavation level
Vibroflotation area Area without
vibroflotation
Figure 73. Results of roller-integrated continuous compaction
control (CCC) on a 1.5 m thick fill on top of subsoil
improved by deep vibroflotation. The positions of the stone
columns are still clearly visible on the CCC diagram.
Omega dimensionless value of compaction degree or soil
stiffness respectively
378 Brandl
Geosynthetics International, 2011, 18, No. 6
geosynthetic; also, local overcompaction of the fill layers
(and hence stress constraints in the geosynthetic inclusions
or cover) is avoided.
Intensive and uniform compaction of geosynthetic-
reinforced dykes, dams, barrier fills and other retaining
structures improves their earthquake resistance signifi-
cantly. If the layers are not properly compacted they will
be more damaged during stronger seismic events. A rigid
facing of geosynthetic-reinforced retaining walls makes
higher compaction of the outer fill zone possible. Conse-
quently, such structures suffer less damage. Over-intensive
compaction, should be avoided, though, because it creates
excessive lateral earth pressures behind rigid facings, or
moves flexible elements outwards. CCC is an excellent
tool for optimising compaction intensity, depending on the
structural requirements.
The increase in lifetime and serviceability of structures
is especially important for road and highway pavements
and for railways, but also for dykes and dams, and for
barrier fills and other retaining structures. Weak spots in
dykes or dams may promote seepage, inner erosion or
piping: hence not only the absolute degree of compaction
but also its uniformity is essential. This can be best
checked by CCC, which is a significant advance over the
statistical quality control used hitherto. Until now the
various selection procedures for spot checking have been
grid pattern, random selection, subjective selection, and
subjective selection using existing criteria.
The CCC method also involves new statistical criteria,
because it records the uniformity of the compacted layer
as well as the increase in degree of compaction during
subsequent roller passes. The relevant parameters are the
mean value, maximum and minimum values, standard
deviation, and increase of the compaction values. More-
over, the minimum quantile can be used as soon as
sufficient experience with CCC is gained on the particular
construction site.
The statistical parameters used hitherto do not allow
assessment of the distribution of control data within a
section: The plots in Figure 74 have the same mean
maximum and minimum values, and standard deviation,
but they exhibit different qualities. Figure 74a shows only
one limited weak zone, which can easily be improved,
whereas the area in Figure 74b requires comprehensive
measures to achieve sufficient and homogeneous quality.
To judge the distribution of control values statistically
within a defined area, the statistical methods of vario-
graphy and kriging can be used. These methodologies
were originally developed in order to interpret geophysical
data, and seem to be a useful tool to improve the quality
assurance of compacted layers, and hence also of geosyn-
theticsoil structures.
A reliable interpretation of the CCC data is possible
only if the operating conditions of the vibratory roller
drum are taken into consideration (see RVS 8.S.02.6 (FSV
2005)). The significant operating conditions depend on
the roller and soil data, and on the interaction between
roller and soil (or other granular material) (Adam 1996;
Brandl and Adam 1997).
Roller compaction technologies have been improved
significantly over the last 15 years, and they now provide
a wide range of possibilities for selecting the most suitable
roller for a particular purpose. A further development is
the automatically controlled VARIOCONTROL roller,
whereby the direction of excitation is controlled automati-
cally by using defined control criteria. VARIOCONTROL
compaction provides uniform compaction, fewer roller
passes, improved compaction in both deeper layers and on
the surface, and reduction of lateral vibrations, for exam-
ple when operating close to sensitive structures.
Smaller compaction equipment (plate vibrators, etc.)
also exists that facilitates CCC for backfills under con-
fined site conditions, or for slender geosyntheticsoil
structures.
To sum up, roller-integrated CCC represents a signifi-
cant improvement for high-quality management systems.
Compaction control is integrated in the compaction pro-
cess, and data are provided over all the compacted area.
Because of the outstanding advantages of CCC, this
technology should be used for the compaction of soil
structures as much as possible, especially in the case of
geosynthetic reinforcement.
To a certain extent, compaction control can be also
achieved by the spectral analysis surface wave method
(SASW) or continuous surface wave technique (CSW).
(a)
(b)
Figure 74. Control plots of compacted areas with the same
conventional statistical characteristics of compaction
(schematic). The strips indicate the roller lanes, controlled
with CCC. The different shadings show different compaction
degrees. Statistical evaluation can be improved with
variography or kriging
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 379
Geosynthetics International, 2011, 18, No. 6
Both methods are non-intrusive, but continuous compac-
tion optimisation (i.e. during rolling) is not possible.
Under confined site conditions, and in narrow places,
dynamic load plate tests (lightweight falling devices) have
proved most suitable. They allow only spot checking, but
many tests can be performed within a short time, and the
results are available immediately: thus reliable quality
assessment is possible.
6. THERMO-ACTIVE GEOSYNTHETICS
FOR TUNNELS (ENERGY
GEOSYNTHETICS)
6.1. General
Natural disasters are increasingly being caused by climate
change. Consequently, measures against climate change
help to mitigate natural disasters. Thermo-active ground
structures represent such a contribution to environmental
protection, and they also provide substantial long-term
cost savings and minimise maintenance (Brandl 2006b).
Energy foundations and energy tunnels make use of earth-
contact concrete elements that are already required for
structural reasons, but simultaneously they work as heat
exchangers. Absorber pipes filled with a heat carrier fluid
are installed within conventional structural elements (piles,
barrettes, diaphragm walls, basement slabs or walls, tunnel
linings), forming the primary circuit of a geothermal
energy system. The natural ground temperature is used as
a heat source in winter, and for cooling in summer: hence
no additional elements have to be installed below the
surface. The primary circuit is then connected via a heat
pump to a secondary circuit within the building.
6.2. Energy tunnels with geocomposites
Energy tunnels are infrastructure tunnels that are used
simultaneously for heating or cooling of buildings, road
pavements, or bridge decks. Until recently geothermal
heating from tunnels was used only in connection with hot
water, mostly without heat pumps, but the heat potential
along a tunnel can also be utilised by using the tunnel
support and lining as energy absorbers. These may be
anchors, rock or soil nails, geosynthetics, or secondary
concrete linings. Anchors or nails reaching deep into the
surrounding ground can activate a relatively large mass for
geothermal utilisation. Energy tunnels may be excavated
as closed systems, for example by the New Austrian
Tunnelling Method (NATM; Figure 75), as TBM-driven
tunnels (with segmental lining), or by the cut and cover
method (open system).
Near the portals of transportation tunnels with geother-
mal equipment the following groups may take the avail-
able energy
the owner or operator of the tunnel
private users (especially large residential blocks, but
also one-family houses)
commercial and industrial users
public users (municipal, federal).
Moreover, seasonal operation makes cooling during
summer possible by using the ground for heat storage via
the energy geocomposites.
An example from a railway tunnel in Vienna underlines
these advantages: about 1200 private flats could be
supplied with geothermal energy, as well as large public
buildings.
Energy tunnels are an exciting challenge to geotechnical
engineering, whereby the optimisation of energy extraction
or feed and storage, of transfer and of distribution requires
a multidisciplinary cooperation. Ground investigation and
geotechnical design should incorporate geothermal aspects
at an early stage. The main advantages of this innovative
technology are as follows.
Commonly, tunnels are situated at a depth where the
seasonal ground temperature is widely constant.
Tunnels exhibit large interfaces between the structure
and the ground, and thus favour the extraction or
feed, and hence storage, of geothermal energy.
Very deep-seated mountain tunnels can make use of
large geothermal gradients.
In long tunnels, significant inner heat is available,
mainly due to the waste heat of transportation. In
metro tunnels, for instance, temperatures of more
than +208C are possible even during the winter
months.
Utilising clean and self-renewable energy from
tunnels is environmentally friendly and economical.
Therefore energy tunnels have high public accep-
tance and political support, which makes the
approval procedures easier.
Until now, mainly cut-and-cover tunnels and open-face
tunnels (excavated with the NATM) have been equipped
Energy
anchors
Primary lining
(shotcrete)
Secondary lining
(shotcrete)
Absorber pipes within
road/railway structure
Energy
geocomposite
Figure 75. Absorber elements for energy tunnels
380 Brandl
Geosynthetics International, 2011, 18, No. 6
with thermo-active geosynthetics or anchors/nails, but
bored tunnels with segmental linings can also be used as
energy tunnels, as they embody earth-contact structural
elements. Furthermore, optimised energy tunnels may use
not only the ground temperature from their large under-
ground contact area, but also inner heat sources from
traffic, lighting etc.
A supply pipe (e.g. along the side walls) feeds those
cellular geocomposites where the absorber fluid takes the
energy from the surrounding ground. The warmed fluid is
then transported from the absorber pipes via collector
lines (HDPE pipes) to a heat pump. Prefabricated geocom-
posites with integrated absorber pipes make possible the
installation of independent thermo-active cells, which can
be easily monitored and separated from the overall circuit
for repair. The geocomposites should be prefabricated in
plant and transported in rolls to the construction site. The
rolls should have a standardised width (typically about
2.5 m; Figure 76) and a length adapted to the cross-section
of the tunnel. The installation procedure in the tunnel is
similar to the placement of conventional nonwoven geo-
textiles. The energy geocomposite has adaptors on both
ends, which have to be overlapped in the longitudinal
direction of the tunnel. The only additional works are
fixing absorber pipes when casting the concrete, and
checking the watertightness of the pipe system.
Absorber pipe systems have to be air-free. Degassing is
therefore essential for heat transfer, and to obtain high
efficiency. Details of installation and degassing can be
found in Markiewicz (2004).
6.3. Large-scale tests in a test hall
Large-scale tests were performed in a geosynthetics
production hall prior to the installation of energy geosyn-
thetics in tunnels.
Since 2002, various geosynthetics have been tested at
the Institute for Soil Mechanics and Geotechnical Engi-
neering of the Vienna University of Technology for
utilisation as geothermal absorber elements in shotcrete-
supported open-face tunnelling (NATM). According to the
current state of the art, nonwoven geotextiles serving
simultaneously for mechanical protection and as in-plane
drainage have proved most suitable. Nonwoven geotextiles
are required in tunnelling mainly for watertight geo-
membrane protection, and they improve roof and sidewall
drainage. Fixing plastic pipes between two geotextiles
provides a thermo-active absorber system with a closed
circuit for energy extraction (or storage).
6.4. Installation of thermo-active geocomposites on
construction site: full-scale tests
In principle, energy geocomposites are installed in a
tunnel in the same way as nonwoven geotextiles, but
special attention must be paid to the connecting pipes
leading from the individual heat exchangers (absorber
pipes) within the geocomposites through the secondary
tunnel lining to the main collector. This detail varies
according to the tunnel structure, being influenced
mainly by the waterproofing system: umbrella or full-
circumference waterproofing, waterproofing with geo-
membrane, or watertight secondary lining (Figure 77). In
the case of groundwater under high pressure, special pipe
passage elements are required.
In order to optimise the production and installation of
thermo-active geocomposites (energy geotextiles) full-
scale tests have performed. The tunnel was excavated in
fractured to intact rock (flysch); the overburden in the test
section varied between 15.6 and 17.3 m. Figure 78 shows
a cross-section of this single-tube railway tunnel, close to
the test fields. The niche allowed the placement of a heat
pump, whereas the test fields extended along the standard
tunnel cross-section without niche. Figure 77 illustrates
the scheme with watertight secondary lining: hence no
waterproofing geomembrane was needed, only a sliding
geomembrane between the thermo-active geotextile and
the watertight concrete.
The following structural elements were used
nonwoven polypropylene geotextile with the techni-
cal data listed in Table 4
strips 2.5 m 3 16.0 m, for which 1.0 m end zones
and 0.5 m wide longitudinal zones were not fitted
with absorber pipes; the longitudinal zones served
Figure 76. Unrolling of an energy geotextile (geocomposites)
Shotcrete
Sliding
membrane
Energy
geotextile
Watertight
inner linning
Bottom
Manifold
Figure 77. Scheme for installing energy geotextiles in a
watertight inner lining
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 381
Geosynthetics International, 2011, 18, No. 6
for fixing the absorber pipes and for overlapping the
geotextiles
absorber pipes of linear polyethylene with copolymer
octane, outer diameter 25 mm, wall thickness
3.5 mm, exhibiting high flexibility and stability,
suitability for bending with small radius but small
rebound forces, and stable long-term behaviour.
The geotextiles were fixed to the primary tunnel lining
(shotcrete) with common nails and, additionally, the
absorber pipes were fixed with pipe clamps to avoid
deformation during concreting of the secondary lining
(Figure 79). Four strips of energy geocomposite were
connected to one thermo-active unit with one manifold for
absorber entry and another one for absorber return. There-
fore the main attention during filling the absorber system
with antifreeze must be paid to complete filling and
degassing of all absorber pipes. Preliminary tests had been
conducted to find a suitable technique for filling the
absorber system, as there is no possibility for degassing at
the top(s) of the absorber system. The passage of the
collecting absorber pipes requires special measures, as
illustrated in the examples of Figures 80 and 81. In
principle, there are four options: plastic screws, flexible
hose, niches, and reinforcement crossing.
The pipes have to withstand the outer pressure of
casting the concrete: therefore they must be kept under
inner pressure (compressed air, p
i
> 2.0 bar) during con-
creting. The formwork transport equipment for the sec-
ondary tunnel lining needs special openings for the pipe
passages. The absorber pipes coming from the individual
geosynthetic strips have to be connected to the collecting
pipes by special T-shaped elements (Figure 82). The
collector pipes should run as far as possible between the
inner and outer reinforcement of the tunnel lining (Figure
83). After finishing all welding work, the entire pipe
system has to undergo detailed tightness testing (Figure
84). The tightness of the pipe system must be checked
after each construction phase: that is, before placing the
reinforcement, after welding the bars, etc.
6.5. Test results
The energy extracted from the ground was transferred to a
heat pump and then to a radiator that emitted the produced
heat to the air in the tunnel. Thus the efficiency of the
thermo-active system could be measured. For parametric
studies, different operating features were investigated. The
measurements comprised (Markiewicz 2004)
16.00 m
Shotcrete
Inner lining
Sliding
membrane
1
.
8
4
m
1
.
8
4

m
9.94 m
7.00
Absorber pipes Absorber pipes
Energy geotextile
Energy geotextile
Watertight connection
Manifold to heat pump
2%
Figure 78. Cross-section of the test site
Figure 79. Installation of energy geotextiles at the test plant
LT22-Bierhauselberg. Pipe loops clearly visible
Table 4. Technical data of the energy geotextile
Property Value
Mass per unit area (g/m
2
) 285
Tensile strength (kN/m) 21.5
Elongation at maximum load (%) 100 (MD)
40 (CD)
Static puncture resistance (CBR test) (N) 3300
Cone drop test (hole diameter) (mm) 17
Permeability (vertical) (l/m
2
s) 70 l/m
2
s
Opening size, O
90
(m) 95
Thickness (2 kPa) (mm) 2.5
382 Brandl
Geosynthetics International, 2011, 18, No. 6
the temperature of the brine solution fluid in the
absorbers (entry, exit)
the temperature of ground/groundwater and tunnel
(air, lining)
the heat production of the radiator
the flow velocities in the heating circuit and absorber
fluid circuit
the electric current consumption of the heat pump
and circulating pumps
photographic documentation of the surface of the
secondary tunnel lining, with thermal imaging
Inner lining
Inner lining
To energy
geotextile
To energy
geotextile
To heat
pump
To heat
pump
(a)
(b)
Figure 80. Watertight passage of the collecting absorber
pipes through inner lining with (a) screws or (b) flexible hose
Figure 81. Pipe passage through the secondary (inner) tunnel
lining. Detail with criss-cross reinforcement
Figure 82. T-shaped elements to connect the absorber pipe
from the individual strips of energy geosynthetics to the
collector pipes
Figure 83. Collector pipe running between the inner and
outer reinforcement of the tunnel lining
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 383
Geosynthetics International, 2011, 18, No. 6
cameras to check the operational serviceability of the
individual strips of the energy geocomposite.
Immediately after the thermo-active operation had
started, the temperatures of the secondary lining changed.
This confirmed an excellent contact between the energy
geotextile and the waterproof concrete.
In the first operating phase, the entry temperature of the
absorber fluid circuit dropped to 18C owing to over-
capacity of the heat pump. The temperature difference
between entry and exit was 3.58C, and the entry tempera-
ture of the heating circuit increased to 218C. Therefore a
heat capacity of only 31.4 W/m
2
could be achieved,
corresponding to a performance factor 4.5. In order to
avoid the formation of ice lenses in the ground, the
capacity of the heat pump was then reduced.
The measurements have shown that a heat capacity of
15 to 20 W/m
2
can be gained from the thermo-active
geocomposites, whereby a significant portion of the heat
comes from the tunnel air. Function tests with a thermal
imaging camera showed that two rings of the energy
tunnel had a reduced capacity (Figure 85). This was
caused by incomplete degassing of an absorber pipe
section.
6.6. Recommendations for practice and further
innovations
Energy geocomposites consisting of nonwoven geotextiles
fitted with absorber pipes can be installed like conven-
tional geotextiles. It is not necessary to use absorber pipes
of such high flexibility as in the large-scale test. Conven-
tional plastic pipes are sufficient if the allowable mini-
mum bending radius is considered.
Fusion welding of the absorber pipes may cause the
formation of bulges inside the pipe seam, thus
reducing the fluid discharge area and eventually the
efficiency. Therefore pipe sockets (e.g. electric
welding sockets) should be preferred.
Conventional attachment of the thermo-active geo-
composite to the primary tunnel lining (shotcrete) is
sufficient. Additional fixing of the absorber pipes by
means of pipe clamps is not necessary.
The number of pipe passages through the
secondary lining (waterproof concrete) should be
minimised. The collecting pipes should not be
placed along the sidewalls of the tunnel, but rather
in the base concrete of the invert. At a central
point (e.g. an emergency exit) they can then be led
outside.
Commonly, tunnel structures with waterproof concrete
consist of the following elements
primary lining (shotcrete)
structured geomembrane with naps (for drainage)
sliding membrane (to minimise transfer of shear
forces)
secondary lining ( waterproof structural concrete).
Figure 84. Tightness check after finishing all welding work
on the pipe system
Section with
energy geotextiles
* 12.0C
12.0
11.0
10.0
9.0
8.0
7.0
6.0
5.0
* 5.0C
(a)
(b)
Figure 85. Pictures of the test site taken with (a) a normal
camera and (b) a thermal imaging camera. Collector pipes
along the side walls of the tunnel are clearly visible in (a).
The brighter area in (b) indicates a test zone with a spot of
lower thermal activity
384 Brandl
Geosynthetics International, 2011, 18, No. 6
With regard to thermo-active geocomposites, the fol-
lowing innovative structure was designed: primary lining
energy geocomposite secondary lining. This geocom-
posite consists of three elements: a nonwoven geotextile,
absorber pipes and a sliding membrane. Geotextiles with a
high in-plane drainage capacity (high transmissivity) can
fulfil the drainage function of a structured geomembrane,
if only moderate water ingress is expected. Figure 86
shows an installation detail for such energy geocompo-
sites.
Sliding membranes of 0.2 mm thickness, commonly
used on construction sites, have proved suitable. Small
leaks would not affect their serviceability, because they
serve not for waterproofing but for separation, with a
smooth interface between geocomposite and waterproof
concrete. This reduces stress constraints within the tunnel
lining, and minimises the transfer of shear forces on the
waterproof concrete (according to the white tank tech-
nology).
6.7. Summary
Thermo-active ground structures (energy foundations,
retaining walls, tunnels, etc.) and also energy wells are
a promising innovation in sustainable and clean energy
consumption. A significant advantage of such systems
is that they are installed within elements that are
already needed for statical/structural or geotechnical
reasons: hence no separate or additional structural or
hydraulic measures are required. Foundations, walls
(below and above ground) or tunnel linings can be used
directly for the installation of absorber pipes for heat
exchange.
Energy tunnels using geosynthetics fitted with absorber
pipes are a key improvement over conventional geother-
mal methods such as deep borehole heat exchangers or
near-surface earth collector systems. Comparative large-
scale tests and site experience have shown that compo-
sites, consisting of nonwoven geotextiles fitted with
absorber pipes provide the best results. Usually, a tempera-
ture difference of only T 28C between absorber fluid
inflow and return flow from the primary circuit is
sufficient for an economical operation of the energy
system. Consequently, such geothermal systems represent
low-temperature systems. Experience has shown that the
electricity required for operating the entire system com-
monly varies between 20% and 30% of the total energy
output. If no heat pump is necessary (e.g. for free cooling)
this value drops to 13% for merely operating a circula-
tion pump.
7. FURTHER APPLICATIONS OF
GEOSYNTHETICS
In addition to the previous chapters some special applica-
tions of geosynthetics are mentioned below
7.1. Energy piles
Energy piles represent an innovative, environmentally
friendly technology for heating or cooling buildings
(Brandl 1998, 2006b). In Austria, nearly 70 000 energy
piles have been installed since the 1980s. They provide
substantial long-term cost savings and minimised mainte-
nance. If piles already required for geotechnical or structur-
al reasons, are equipped with absorber pipes, they work
simultaneously as bearing elements and as heat exchangers.
Moreover, they may be combined with diaphragm walls,
rafts or basement walls (e.g. Figure 87). However, in very
soft soil the diameter of in situ cast concrete piles may be
constricted. This can be avoided by wrapping the reinforce-
ment cage of the piles with highly stretchable nonwoven
geotextiles (welded). Bulging can be minimised as well.
When casting the fresh concrete, the lateral pressure of the
fluid stretches the geotextile, so that a sufficient concrete
cover of the reinforcement is attained. This technique
20 20 20
40
Absorber pipes
25 mm dia.
Sliding
geomembrane
Nonwoven
geotextile
Figure 86. Installation detail for overlapping thermo-active
energy composites: Nonwoven geotextile (with high
transmissivity), absorber pipes, sliding geomembrane
Thermo-active
concrete slabs
Secondary
thermal circuit
Thermo-active
concrete slabs
Energy piles
Primary
thermal circuit
E
n
e
r
g
y

d
i
a
p
h
r
a
g
m

w
a
l
l
s
Figure 87. Cross-section through a building with geothermal
cooling and heating; energy piles and energy diaphragm
walls in very soft ground
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 385
Geosynthetics International, 2011, 18, No. 6
provides the required pile integrity, and hence the designed
heat transfer from and into the ground.
7.2. Rockfall shelters
Roads and railways along steep rock slopes need shelters
if protective fences are not sufficient to withstand severe
impacts. Large-scale blocks of rock falling from great
height may destroy even prestressed reinforced concrete
structures if they hit them directly (Figure 88). Therefore
shelters under high dynamic load should be covered by
sloped soil fill, preferably reinforced with geosynthetics
(Figure 89): thus the blocks can roll over the structure
without damage.
7.3. Sandwich reinforcement of embankments and
barrier dams
The stability and deformation behaviour of high embank-
ments and barrier dams along steep or unstable slopes, or
in seismic areas, can be significantly improved by geosyn-
thetic sandwich reinforcement. The scheme in Figure 90
shows that the reinforced zones should preferably be
placed at levels where benches are designed. The top
reinforcement is recommended if the embankment carries
traffic routes. It serves for better load distribution, and
reduces differential settlements. Moreover, sandwich rein-
forcement provides a high earthquake resistance.
7.4. Glacier melting
Glacier melting progressively creates unstable slopes,
when ground that had been frozen into great depth is
thawing. High pore-water pressures and seepage reduce
local and global slope stability, and trigger debris flows,
multiple slips or deep-reaching block sliding. Moreover,
glacier melting becomes a problem in tourist regions,
especially where Alpine skiing dominates. Therefore
several precautionary measures and contingency plans
have been developed in Austria. A special application of
geosynthetics is the local cover of exposed glacier areas
and zones around lift columns with white, nonwoven
geotextiles (Figure 91). Lift columns are frequently
founded or anchored in creeping ice that undergoes in-
creasing differential movements. A local reduction of
glacier melting by geotextile cover prolongs the intervals
of column readjustment.
Several in situ tests were performed to find a method to
reduce glacier melting (Figure 92). Measurements showed
that special geotextiles provided the optimal results, with
a reduction of about 60%. The main characteristics of the
most suitable protective geosynthetics (found from com-
parative in-situ tests) are listed in Table 5.
7.5. Beaver barriers
An increasing population of beavers also represents an
increasing danger to dykes and flood protection dams.
Therefore field tests with several geosynthetics as beaver
barriers were conducted in cooperation with zoologists
(Figure 93). The tests showed that the beavers eventually
broke through all the geosynthetics, but not through the
metal wire mesh with synthetic cover. Such wire meshes
have been installed in Austria since 2008 along critical
sections through and beneath dykes and dams. At the toe
of the dam they are fixed in a trench filled with concrete
Figure 88. Concrete shelter damaged by heavy rockfall
O
r
i
g
i
n
a
l

t
e
r
r
a
i
n
O
r
i
g
i
n
a
l

t
e
r
r
a
i
n
Anchor wall
(on filter concrete)
Reinforced with
geosynthetics
Final
terrain
2
:
3
Sockets Dolomite
(mylonitic)
A
l
r
A
1000 kN
2535 m

Prestressed
anchors
Variable
length
1:5
2
:
3
140 m
4
.
7

m
Figure 89. Reinforced concrete shelter (gallery) against
rockfall, debris flow and avalanches, covered by protective fill
reinforced with geosynthetics; A
r
is anchor force
Interlocking
(toothed)
Top reinforcement for
roads, highways or
heavy buildings
(Possibly interlayers
in between)
Reinforced interlayers
( 36 m) h
Not reinforced
Natural ground
(unstable)
Figure 90. Sandwich reinforcement of high embankments
and barrier dams along steep or unstable slopes and in
seismic zones. Scheme with steepened fill slope; not to scale
386 Brandl
Geosynthetics International, 2011, 18, No. 6
(Figure 94); at the crest they are covered by the layers of
the road structure.
7.6. Geosynthetics for in-situ groundwater cleaning
Permeable reactive walls or funnel and gate systems have
proved suitable for the in situ cleaning of contaminated
groundwater (Figure 95). The contaminant plume then
flows through a straight or curved wall, or is directed to a
gate. Groundwater cleaning in the reactive wall or gate is
performed site-specifically, whereby physical, chemical,
and microbiological measures are possible. Several sys-
tems contain exchangeable geosynthetic filter panels, and
also geotextiles to envelope special (granular) reactive
material.
Figure 91. Local geotextile cover of glaciers to reduce
melting of particular zones
Without measures
TenCate Topex GLS
Water injection
Snow compaction
300
250
200
150
100
50 M
o
l
t
e
n

t
h
i
c
k
n
e
s
s

o
f
i
c
e
/
s
n
o
w

(
c
m
)
60% reduction
of snow melting
July 2005 September 2005
Time
Figure 92. Reduction of glacier melting by placing a special
nonwoven geotextile cover; results of in situ tests
(a)
(b)
(c)
Figure 93. (a, b) Geosynthetics and (c) wire mesh to prevent
damage to flood protection dams caused by beavers. Field
tests in cooperation with zoologists
Table 5. Geotextile data to reduce glacier melting
Property Value
Geotextile Nonwoven
polypropylene
Mass per unit area (g/m
2
) 340
Thickness (at 2 kPa) (mm) 2.0
UV resistance (residual strength) (%) .75
Tensile strength (EN ISO 10319) (kN/m) 23
Opening size (EN ISO 10319) (mm) 0.06
Thermal resistance (m
2
K/W) 0.0593
Water permeability in plane (m
3
/ms) 2 310
6
Water penetration resistance (cm) .8
Light reflection capacity (wave length 500 mm) (%) 78
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 387
Geosynthetics International, 2011, 18, No. 6
7.7. Contaminant-absorbent geosynthetics
Contaminant-absorbent geosynthetics are not only used in
reactive walls to clean groundwater but also for surface
measures. In both cases nonwoven geotextiles are pre-
ferred for environmental protection. Whereas reactive
walls are installed more for long-term purposes, surface
measures serve for emergency action (Figures 96 and 97).
Absorbent geosynthetics used for surface protection or
cleaning may be socks, pillows, mats, rolls or oil booms.
Furthermore, oil skimmers may absorb oil and fuel in
canals, bilges, sumps, manholes, drains and tanks.
Absorbent nonwoven geotextiles consist of polypropy-
lene. Their costs are clearly below those of granular
absorbers, and their absorbent capacity is significantly
higher. The used (and hence contaminated) absorbent
geosynthetics are most suitable for waste incineration and
thermal utilisation. The main characteristics of the hitherto
optimised product are
production technology: meltblown
fibre fineness: 58 m
bonding type: thermally calender or ultrasonic
bonding
engraving: point bonded, flat bonded
mass per unit area: 15400 g/m
2
.
Oil booms and skimmers consist of polypropylene
geotextile strips covered by a nylon net of high tensile
strength.
Figure 94. Installation of nets against beaver activity along
flood protection dams
Ground plans
Contamination
source
Contamination
source
Reactive wall
Cleaned
groundwater
Cleaned
groundwater
Cleaned
groundwater
Cleaned
groundwater
Contamination plume Contamination plume
GW flow direction
GW flow direction
Funnel
Unsaturated zone
Unsaturated zone
Aquifer
Aquifer
Aquitard Aquitard
k k
f r
-Wall -Aquifer
(a)
k k
k k
r f
r f
-Gate -Aquifer
-Funnel -Aquifer

(b)
Gate
Cross-sections
Figure 95. In situ groundwater cleaning with (a) permeable reactive walls or (b) funnel and gate system
Figure 96. Placing of geosynthetic mats to absorb
contaminating liquids on a river
388 Brandl
Geosynthetics International, 2011, 18, No. 6
8. FINAL REMARKS AND
CONCLUSIONS
Geosynthetics engineering offers a broad range of
applications for damage prevention, natural disaster
mitigation, rehabilitation measures, and environmental
protection. Coastal protection, erosion protection, seismic
aspects, sinkhole bridging and hazardous waste contain-
ment, for instance, are topics that have been treated in
numerous publications. Therefore they are not discussed
in this Giroud Lecture, although referred to in most
sections.
The selected geosynthetic applications emphasize the
close links between theory and practice, and between
geosynthetics and geotechnical engineering. Design by
function, or semi-empirical design with calculated risk
(interactive design), combined with the observational
method, represents an essential feature of geosynthetics
application. For geosyntheticsoil systems in critical
areas (unstable slopes, seismic zones, etc.) contingency
plans should exist (e. g. possibilities of in-time strength-
ening).
Because of their flexibility, geosyntheticsoil struc-
tures have clear advantages over rigid structures with
classical construction materials: higher impact absorp-
tion of rockfalls or avalanches, lower sensitivity to
earthquake and creeping slopes, and rapid installation
(for flood defence, etc.). Ecological advantages can also
be found when comparing the total energy consumption
for the component and final products, for transportation
from the manufacturer to the construction site, and for
installation.
ACKNOWLEDGEMENTS
This lecture was presented at the 9th International Con-
ference on Geosynthetics, in 2010, in Guaruja, Brazil. The
author wishes to thank the publishers (Brazilian Chapter
of the International Geosynthetics Society) and the organi-
zers (E. M. Palmeira, D. M. Vidal, A. S. J. F. Sayao and
M. Ehrlich) for permission to publish this paper in an
updated and extended version.
NOTATIONS
Basic SI units are given in parentheses.
A area (m
2
)
A9 corrected area (m
2
)
c cohesion (N/m
2
)
c
r
residual cohesion (at larger shear deformations)
(N/m
2
)
D
Pr
degree of compaction (percentage of Proctor
density)
E
cr
creep pressure (N/m
2
)
E
kin
kinetic energy (J)
E
pot
potential energy (J)
E
v
deformation modulus (N/m
2
)
F safety factor (dimensionless)
F horizontal shear force (N)
FS factor of safety (dimensionless)
G shear modulus (N/m
2
)
H height (m)
h height (m)
i
crit
critical hydraulic gradient (dimensionless)
ss self-settlement (mm)
performance factor (dimensionless)
slope angle (degrees)
r density (g/m
3
)
friction angle (degrees)

r
residual shear angle (degrees)

design
design shear angle (degrees)

peak
peak value in the curve versus shear
deformation (degrees)
REFERENCES
Adam, D. (1996). Continuous Compaction Control with Vibratory
Rollers. Doctoral thesis, Vienna University of Technology, Vienna,
Austria (in German).
Akkerman, G. J., Bernardini, P., van der Meer, J., Verheij, H. & van
Hoven, A. (2007). Field tests on sea defences subject to wave
overtopping. Coastal Structures 2007: Proceedings of the 5th
Coastal Structures International Conference, CSt07-2, Venice, Italy,
Franco, L., Tomasicchio, G. R. & Lamberti, A., Editors, World
Scientific, Singapore, vol. 1, pp. 657668.
Blovsky, S. (2002). Bewehrungsmoglichkeiten mit Geokunststoffen.
Doctoral thesis, Institute for Soil Mechanics and Geotechnical
Engineering, Vienna University of Technology, Vienna, Austria (in
German).
Brandl, H. (1979). Design parameters for weak rocks. Panel discussion.
Proceedings of the 7th European Conference on Soil Mechanics
and Foundation Engineering, Brighton, UK, A.A. Balkema,
Rotterdam, the Netherlands, vol. 4, pp. 6367.
Brandl, H. (1980). Theory of earth pressure and creep pressure of sliding
slopes with practical examples. Proceedings of the 6th Danube-
European Conference on Soil Mechanics and Foundation Engineer-
ing, Varna, Bulgaria.
Brandl, H. (1998). Energy piles and diaphragm walls for heat transfer
from and into the ground. Proceedings of the 3rd International
Seminar on Deep Foundations on Bored and Auger Piles, Ghent,
Belgium, Van Impe, W. F., Editor, A.A. Balkema, Rotterdam, the
Netherlands, pp. 3760.
Brandl, H. (2006a). Bridges or embankments in creeping slopes?
Darmstadt Geotechnics, Technical University Darmstadt, Germany,
vol. 11, pp. 7392.
Brandl, H. (2006b). Energy foundations and other thermo-active ground
structures. Rankine Lecture. Geotechnique, 56, No. 1, 79122.
Figure 97. Oil booms of absorbent geosynthetics to contain
and absorb oil spills on water
Geosynthetics applications for the mitigation of natural disasters and for environmental protection 389
Geosynthetics International, 2011, 18, No. 6
Brandl, H. & Adam, D. (1997). Sophisticated continuous compaction
control of soils and granular materials. Proceedings of the 14th
International Conference on Soil Mechanics and Foundation
Engineering, Hamburg, Germany, A.A. Balkema, Rotterdam, the
Netherlands, pp. 3136.
Brandl, H. & Adam, D. (2000). Special applications of geosynthetics in
geotechnical engineering. Keynote Lecture. Eurogeo 2000: Pro-
ceedings of the 2nd European Geosynthetics Conference, Bologna,
Italy, Cancelli, A., Cazzuffi, D. & Soccodato, C., Editors, Patrone
Editore, Bologna, Italy, vol. I, pp. 27 64.
Brandl, H. & Hofmann, R. (2006). Erosionsstabilitat und Standsicherheit
von Schutzdammen gegen Wildbache und Murengange mit
besonderer Berucksichtigung von Einbauten. Sicherung von Dam-
men, Deichen und Stauanlagen, Hermann, R. A., Jensen, J., Editors,
Universitat Siegen, Germany, vol. I., pp. 139171 (in German).
CEN (2004). EN 1997-1: Eurocode 7: Geotechnical Design Part 1:
General Rules. Comite Europeen de Normalisation, Brussels,
Belgium.
Chugaev, R. R. (1965). Calculation of the filter stability of the ground
below dams. Gidrotechniceskoe Stroitelstvo, No. 2 (in Russian).
Descoeudres, F. (1997). Aspects geomecaniques des instabilites de
falaises rocheuses et des chutes de blocs. Mitteilungen der
Schweizerischen Gesellschaft fur Boden und Felsmechanik, Mon-
treux, Switzerland (in French).
Egloffstein, Th. (2009). O

kologischer Vergleich zwischen Bauweisen mit


mineralischen Baustoffen und Bindemitteln sowie Bauweisen mit
Kunststoffen. Geotechnik, Deutsche Gesselschaft fur Geotechnik,
Essen, Germany, vol. 3, pp. 156163.
FSV (2005). RVS 8.S.02.6: Roller-integrated Continuous Compaction
Control (CCC), Austrian Guidelines and Codes for Roads, Austrian
Association for Research on Road - Rail - Transport (FSV), Vienna,
Austria.
Giroud, J. P. (2003). Filter criteria. Mitteilungen fur Grundbau,
Bodenmechanik und Felsbau, Technische Universitat Wien, Vienna,
Austria, No. 5 (Jubilee volume), pp. 221259.
Giroud, J. P. (2010). Development of ciriteria for geotextiles and granular
filters. Prestigious Lecture 1. Proceedings of the 9th International
Conference on Geosynthetics, Guaruja, Brazil, pp. 4566.
Gutknecht, D., Reszler, Ch. & Bloschl, G. (2002). Das Katastrophen-
hochwasser vom 7. August 2002 am Kamp eine erste
Einschatzung. Elektrotechnik und Informationstechnik, Jahrgang,
119, 411413.
Heerten, G. (2006). Was hat sich verandert seit der Elbe-Flut 2002?
Sicherung von Dammen, Deichen und Stauanlagen Hermann, R.
A, Jensen, J., Editors, Universitat Siegen, Germany, vol. II, pp.
1935.
Heibaum, M., Fourie, A., Girard, H., Karunararne, G. G., Lafleur, J. &
McGrath, J. (2006). Hydraulic application of geosynthetics. Special
Lecture. Proceedings of the International Conference on Geosyn-
thetics (IGS), Yokohama, Japan, Millpress, Rotterdam, the Nether-
lands, pp. 79120.
Hoek, E. & Gralewska, A. (1969). A Rock Mechanics Information
Service, Rock Mechanics Research Report No. 9, Imperial College
London, UK.
Koerner, R. M. (1998). Designing with Geosynthetics, 4th edition,
Prentice Hall, Englewood Cliffs, NJ, USA.
Lawson, C. R. (2008). Geotextile containment for hydraulic and
environmental engineering. Geosynthetics International, 15, No. 6,
384427.
Markiewicz, R. (2004). Numerische und experimentelle Untersuchun-
gen zur Nutzung von geothermischer Energie mittels erdberuhrter
Bauteile und Neuentwicklungen fur den Tunnelbau. Doctoral
thesis, Institute for Soil Mechanics and Geotechnical Engineer-
ing, Vienna University of Technology, Vienna, Austria (in
German).
Pichler, H. (2003). Parameter Identification (PI) as the Basis for
Prognoses in Geotechnical Engineering. Doctoral thesis, Vienna
University of Technology, Vienna, Austria.
Pitanga, H. N., Gourc, J.-P. & Vilar, O. M. (2009). Interface shear
strength of geosynthetics: evaluation and analysis of inclined plane
tests. Geotextiles and Geomembranes, 27, No. 6, 435446.
Recio, J. & Oumeraci, H. (2008). Hydraulic permeability of structures
made of geotextile sand containers: laboratory tests and conceptual
model. Geotextiles and Geomembranes, 26, No. 6, 473487.
Saathoff, F., Oumeraci, H. & Restall, S. (2007). Australian and German
experience on the use of geotextile containers. Geotextiles and
Geomembranes, 25, No. 46, 251263.
Vanicek, I. & Vanicek, M. (2008). Earth Structures in Transport, Water
and Environmental Engineering, Springer, Dordrecht, the Nether-
lands.
VDZ (2002). Hochwasserschutz und zementgebundene Baustoffe,
Hinweise fur Planung und Ausfuhrung. Bundesverband der
Deutschen Zementindustrie e.V. und des Vereins Deutscher
Zementwerke e.V., Dusseldorf, Germany (in German).
Walz, B. (1982). Bodenmechanische Modelltechnik als Mittel zur
Bemessung von Grundbauwerken. Forschungs- und Arbeitsberichte
aus den Bereichen Grundbau, Bodenmechanik und unterirdisches
Bauen, Bergische Universitat GH Wuppertal, Germany, no. 1 (in
German).
Ziegler, F. (1998). Mechanics of Solids and Fluids. Springer, New York,
NY, USA.
Ziems, J. (1967). Neue Erkenntnisse hinsichtlich der Verformungsbes-
tandigkeit der Lockergesteine gegenuber Wirkungen des Sickerwas-
sers. Wasserwirtschaft Wassertechnik 17, No. 7, 5055 (in
German).
FURTHER READING
Brandl, H. (2001). High embankments instead of bridges and bridge
foundations in embankments. Geotechnics for Roads, Railtracks
and Earth Structures, Correia, A.G., Editor, A.A. Balkema, Lisse,
the Netherlands, pp. 1326.
Heerten, G. & Horlacher, H.-B. (2004). Reconstruction and improvement
of dykes by using local soil and geosynthetics. Eurogeo 3:
Geotechnical Engineering with Geosynthetics, Proceedings of the
3rd European Geosynthetics Conference, Munich, Germany, Floss,
R., Brau, G., Nubaumer, M., Laackmann, K., Editors, Deutsche
Gesselschaft fur Geotechnik, Essen, Germany, vol. II, pp 371376.
Jaecklin, F. P. (2002). Bemessung von kunststoffbewehrten Stutzmauern
und Steilboschungen: Grundlagen, Methoden, Vorgehen. Proc.
Fachtagung des Schweizerischen Verbandes der Geotextilfachleute,
Dubendorf, Zurich, Switzerland (in German).
Koerner, R. M. & Guglielmetti, J. (1995). Vertical barriers: geomem-
branes. Assessment of Barrier Containment Technologies: A
Comprehensive Treatment for Environmental Remediation Applica-
tions, Rumer, R. R. & Mitchell, J. K., Editors, NTIS, US
Department of Commerce, Washington, DC, USA, PB 96-180583,
pp. 95118.
The Editor welcomes discussion on all papers published in Geosynthetics International. Please email your contribution to
discussion@geosynthetics-international.com by 15 June 2012.
390 Brandl
Geosynthetics International, 2011, 18, No. 6

Вам также может понравиться