Project 4: Stability of rock slopes
Project manager: Reginald Hermanns (NGU)
Rock slope failures -
Models and Risk
The project seeks to integrate the
geological and geotechnical/geomechanical aspects of hazard and risk
related to large rock-slope failures.
View towards the slide scar and the
deposits of a large rock avalanche in Balsfjord, Troms, Norway
Project description
Overview
Themes
Recent results
Key personnel
NGU: Lars Blikra (project manager),
Marc-Henri Derron, Ian Henderson,
NGI: Rajinder Bhasin, Ulrik Domaas, Vidar
Kveldsvik (PhD)
NTNU: Bjørn Nilsen, Guro Grøneng (PhD)
NORSAR: Michael Roth
University of Bergen/NGU: Alvar Braathen
International
corporations
Geological Survey of Canada
ETH-Zürich
University of Milano
A large rock-slope
failure at top of the Oppstadhornet mountain, west of Molde in Møre &
Romsdal. The slide involves a volume of more than 20 million m3. Parts
of the slide are active, and if the failure develops into a rock
avalanche it will create damaging tsunamis.
Overview of the
project
The
project will focus on risk related to large rock-slope failures. Rock
avalanches and related tsunamis represent one of the most serious
natural hazards in Norway, and during the last 100 years more than 170
people have lost their lives only in northern West Norway. Also on a
global basis, such events have caused major disasters.
The project seeks to
integrate geological and geotechnical/geomechanical aspects of geohazard
related to rock-slope failures, and is organized as a corporation
between all the ICG partners. Two PhD students are integrated in the
project, focusing on geological models, numerical modelling and
kinematics of rock-slope failures. The projects so far have focused on
the study of rock-slope failures in western and northern Norway, and
include several aspects related to rock-slope failures and stability (Blikra
et al., 2004; Braathen et al., 2004; Bhasin & Kaynia, 2004; Bhasin et
al., 2004; Dahle, 2004; Panthi & Nilsen, in press).
The project personnel
is also directly involved in projects related to risk analysis and
investigations and monitoring of high-risk objects in the counties. Much
of the work has been related to
investigation and monitoring at the Åknes failure, were major external
funding will be available. This will include geological and geophysical
investigations, drilling with instrumentation and logging, installation
of different monitoring systems and modelling approach both related to
stability, slide dynamics and tsunamis.
Satellite
data and swath bathymetry from Tafjord, showing the location of
rock-avalanche deposits onshore and in the fjord.
Themes
A. Methods for
quantification of rockslide hazard
The
project use all available data that exists in Møre & Romsdal to evaluate
methods for hazard assessments in a selected fjord area, including
regional and more local hazard and risk zoning (e.g. Blikra et al.,
2004). The projects will both evaluate methods for regional hazard
zoning and for more detailed hazard and risk zg. The reoningional hazard
zonation has been performed based on the spatial distribution and
temporal pattern of events, and several dating methods have been
employed. Further work is required on exposure dating. Secondly, a more
detailed and quantitative zonation will be performed in a selected fjord
area. This is mainly based on the frequency, age and size of events in
defined areas. In addition, the run-out and tsunami potential needs to
be taken into account.
Rock avalanche events and regional hazard
zones in the Møre & Romsdal county, western Norway
Illustration: Shaded
relief model based on swath bathymetry data from Sunnylvsfjorden in Møre
& Romsdal county, showing large rock-avalanche deposits.
The numbers indicate
the volume of individual events in million m3. These type of data
together with reflection seismics are the base for evaluating the hazard
level (distribution, frequency and magnitude of events).
B. Development of geological models for
rock slopes
The most
fundamental input for rock-slope failure assessment comes from the
geological dataset. Hence, realistic models based on in-depth
understanding of structures, 3D kinematics and stability related to
friction (fractures, fault rock membranes) are crucial. The following
topics will be focused on
- Digital
elevation models (DEM). DEM will be used to characterize
structural patterns of importance for instability features (e.g.
sliding planes and wedges). Recent technologies such as the airborne
light detection and ranging (lidar) or the ground-based laser scan
provide digital elevation models (DEM) with a submetric resolution. A
high resolution DEM can be used to provide a first image of the
phenomena: geometry and limits of the instability, main sets of
discontinuities and their orientations, kinematic tests of planar
sliding or wedge failure, and a first estimation of the volumes
involved. These types of methods have already been tested on two
Norwegian sites (Oterøya and Åknes) and new data from the Tafjord area
are now available. This project will focus on two aspects: (1) the
acquisition of ground-based and airborne laser scan data of rock
slopes, and (2) the development of methods for a preliminary
assessment based on a high resolution DEM.
A) orthophoto of the
upper part of the Åknes landslide. The white line is the open fracture
delimiting the top of the landslide. B) Color-DEM representation of
the same area than in A with some slope measurements (dip
direction/dip angle). The color coding is given in the stereographical
projection on the right. C) Identification on the DEM of the cells
having an orientation corresponding to the discontinuities J1, J2, or
J3 (same area than in A); (illustration from Derron et al. in press).
- 3D geometry of
rock-slope failures.
The focus will
be on detailed investigations of slope-failures in order to identify
diagnostic structures for such features. Mapping requires use of
geophysical methods (2D resistivity, seismics) in order to identify
deeper and sole structures. Drilling should also been performed if
possible. Priority will be given to the localities Åkerneset and
Oppstadhornet in Møre & Romsdal and Nordnes in Lyngen, Troms.
A 2D resistivity
profile crossing a large rock-slope failure at Nordnes in Lyngen,
Troms county. In general there are high resisitivity values, except
for a nearly vertical zone showing distinct lower values (blue
colours). This zone represents the major fractures in the upper part
of the sliding mass, demonstrating that the slide is at least 100 m
deep in the upper part..
- Geo-mechanical
uncertainties in rock slope stability assessment.
Present consensus and numerical models for
rock-slope stability, utilised in risk assessments, is based on
shear-fracture characteristics. Ongoing outcrop studies on
collapsing mountain-sides (e.g., Blikra et al. 2004; Braathen et al.
2004; Dahle, 2004) clarify that most rock-slope failure areas
contain both (i) shear - and (ii) tension fractures, as well as a
basal shear surface commonly covered with a membrane of crushed rock
(fault breccia/gouge). The project will
address uncertainties related to variables applied in numerical
rock-slope stability models, including laboratory analysis (e.g.,
ring-shear study). Especially important is to establish the
shear-strength relationships for unconsolidated, water saturated
fault rocks along sliding planes and to establishing frictional
relationships for large-displacement rock fractures in gneiss.
Profile illustrating a
two-layer model for pre-avalanche deformation in a rock-slope failures
area. Text-boxes describe factors an mechanical properties that have
to be evaluated in stability assessments (from Braathen et al., 2004).
C. Stability
analysis, including sensitivity and probabilistic analysis.
Major focus will be
on stability analysis and modelling, and this will be highly
integrated with the work on geological models. Discontinuum numerical
modelling was performed for a 700 m high rock slope in western Norway
(Oppstadhornet), in both static and dynamic rock slope stability
analysis. One of the PhD studies (Vidar Kveldsvik) will be focused on
numerical modelling, sensitivity and probabilistic analysis. The work
will be concentrated on the Åkernes and Oppstadhornet sites.
Especially important will be the analysis of uncertainties in Barton &
Bandis parameters. The collected geotechnical data from the field
will be analysed statistically in order to determine variation
associated with the parameters and assess the appropriate statistical
distributions. In the existing literature the slope stability
analysis by numerical techniques and probabilistic methods are
performed separately. This research, for the very first time, proposes
to combine both the approaches so that they will benefit and
compliment each other and will better explain the phenomenon.
Two master students at NTNU will also be integrated in the ICG project
focusing on rock-slope stability in Nepal. Master students also
focus on the Tafjord area.
Illustration:
Displacement vectors and shear displacements (indicated by thick lines)
of a jointed rock slope after initial static loading
D. Monitoring and
deformation process (3D kinematics).
 The
understanding of the 3D kinematics (movement pattern) is of major
importance for the evaluation of hazard and for numerical modelling. A
PhD student (Guro Grøneng) will focus on creep and deformation
processes, including 3D modelling. The project will use existing
monitoring data, e.g. from the Åknes site in Møre & Romsdal, were
automatic extensometers have been operating more than 10 years. Several
new monitoring systems will be installed at this site during the next 2
years, and these data will be an essential part of these studies.
Movement measured by
the extensometers at the Åknes failure.
E. Microseismic monitoring
Another monitoring
method that was explored at the Åknes failure was passive microseismic
monitoring. Microseismic events that are expected to be seen in
unstable rock slope sites are events directly related to the mass
movement (e.g. shear failures at the detachment plane, opening of
fractures, etc) and second order events associated with rock falls,
small-scale slides etc. The spatial and temporal distribution of
microseismic events can provide valuable information on the internal
structure and the dynamical behaviour of the slope. In a pilot study a
temporary small-scale seismic network consisting of 6 geophones was
installed during the summer/fall 2004. The main purpose of the pilot
installation was to check, if we can see any seismic activity at all
and to get an idea on the noise conditions.
Even with relatively
insensitive geophones we observed more than 350 microseismic events
during the monitoring period of 71 days. We considered signals as
microseismic events, if they had a short duration (less than 5 sec)
and if they could be observed on at least 4 geophones with a
signal-to-noise ratio better than 2. Excluding positively identified
man-made events we end up with a rate of about 3.5 microseismic events
per day. The network was too small and the signal quality too poor to
localize the seismic events accurately, but we could determine the
direction of some of the strongest incoming signals. They are
generated within the unstable part of the slope, downhill from the
geophone network
For 2005 we plan to
install a temporary network with larger aperture and more receivers.
In the framework of the investigations of the Åknes site, boreholes
will be drilled for direct sampling and for the instrumentation with
different sensors. Depending on the progress of the drilling we might
be able to install borehole seismometers at the end of the field
campaign 2005, and set up a permanent network.
Examples of microseismic
events recorded with the temporary network: (a) complex source (probably
a rock fall); (b) event with clear P-and S-wave onsets; (c) ‘typical’
event representative for most of the observed events; (d) map view of
the seismic network and directions of the strongest events
F. Rockslide
dynamics and empirical modelling
Numerous rock avalanche
events have been mapped in Norway, of which some exceed 100 million m3
in volume. Study of rock-avalanche processes, including slide dynamic
and secondary processes related to slide impact, is important for
tsunami modelling and for understanding run-out potential. The Norwegian
rockslide examples will be excellent for developing new statistical
models for run-out distance. The project will study a number of these
sites as a base for development of an empirical run-out model for large
rock avalanches.
View towards the slide scar and the
deposits of a large rock avalanche in Balsfjord, Troms, Norway.
The distinct slide scar demonstrates that the failure
occurred along the foliation planes. The deposits show that the volume
is more than 100 million m3, and with a run-out distance of 5 km.
Recent results
Publications
Anda, E., Blikra, L.H.,
& Braathen, A. 2002: The Berill fault - first evidence of neotectonic
faulting in southern Norway. Norwegian Journal of Geology (NGT) 82,
175-182.
Blikra, L.H., Longva,
O., Braathen, A. & Anda, E., Dehls, J. & Stalsberg, K. 2005: Rock-slope
failures in Norwegian fjord areas: examples, spatial distribution and
temporal pattern. In Evans, S.G., Scarawcia Mugnozza, G., Strom, A.L. &
Hermanns, R.L. (eds.), Massive rock slope failure: new models for hazard
assessment. Kluwer, Dodrecht (in press).
Braathen, A., Blikra,
L.H., Berg, S.S. & Karlsen, F. (2004): Rock-slope failures of Norway;
type, geometry, deformation mechanisms and stability. Norwegian Journal
of Geology (NGT) 84, 67-88.
Bhasin, R. and Kaynia,
A.M. (2004): Static and Dynamic Simulation of a 700 m High Rock Slope in
Western Norway. International Journal of Engineering Geology 71,
213-226.
Bhasin, R., Kaynia,
A., Blikra, L.H., Braathen, A. & Anda, E. (2004): Insights into the
deformation mechanisms of a jointed rock slope subjected to dynamic
loading. International Journal of Rock Mechanics & Mining Sciences. 41
Derron, M.H.,
Jaboyodoff, M, & Blikra, L.H. 2005: Preliminary assessment of rockslide
and rockfall hazards using a DEM (Oppstadhornet, Norway). Natural
Hazards and Earth System Sciences 5, 285-292.
Hermanns, R.L., Blikra,
L.H., Naumann, M., Nilen, B., Panthi, K.K., Stromeyer, D. & Longva, O.
2005: Examples of multiple rock-slope
collapses from Köfels (Ötz valley, Austria)
and western Norway. Engineering Geology (in press).
Panthi K. K. and
Nilsen B. (in press): Numerical Analysis of Stresses and Displacements
for the Tafjord Slide, Norway. Bulletin of Engineering Geology and the
Environment.
Reports
Bhasin, R. 2004: Rock
slope failures – Models and risk. Stability analysis of rock slopes. ICG
Report 2004-4-2
Lindholm, C. 2003:
Site-specific hazard for three sites in Northwestern Norway. ICG Report
Roth, M. 2003:
Report on seismic monitoring of rock slopes. ICG
Report
Dahle, H. 2004:
Analysis of Slope Stability at Opstadhornet (in Norwegian). Master
thesis. Dept. of Geology and Mineral Resources Engineering, NTNU".
Domaas, U. 2004:
Rock-slope failures – Models and Risk.
Monitoring systems: ICG Report 20031093-1 |
