Monitoring evolution processes of future rockslides

Rockslides exact a heavy social, economic and environmental toll in populated mountain areas. Recent rockslides in the Swiss Alps demonstrate the need for a deeper understanding of the geological and physical processes that lead to catastrophic slope failure. ETH Zürich has initiated a project aimed at studying evolutionary failure processes that result in large-scale mass movements in massive hard/brittle rock slopes. The working hypothesis of the project contends that rock-slope instability occurs via progressive development of a failure plane along a path that passes through existing discontinuities, either in a planar or stepped-path fashion. This development involves a complex process of fracture nucleation, propagation and coalescence, combined with spatial and temporal variations in pore pressures. Various techniques are employed to image potential failure planes and to monitor their dynamic behavior. Spatial characterization is achieved with geological mapping and tomographic inversions of seismic and georadar data collected at the surface and in boreholes. Studying the dynamic behavior requires a monitoring system to be installed. Data collected with such a monitoring system can provide, for example, information on the locations of actual movements, displacement rates and deformation mechanisms.

After extensive tests at several sites in the Swiss Alps, the area of the Randa rockslide, located in the western Swiss Alps, was identified to be most suitable for our purposes (Figure 1). Here, approx. 30 millions m3 of rock material was released during a large rockslide in 1991. Geodetic measurements performed during the past 12 years indicate ongoing surface displacements of about 10 mm per year. It is assumed that these displacements manifest themselves in the form of small-scale microseismic events along existing and/or newly developing fractures.

Our primary investigation area lies on top of the rockslide (Figures 1 and 2). Here, three boreholes were drilled to depths of 50 - 120 m. They were equipped with deformation measurement devices (inclinometers, extensometers, TDR cables). At the bottom of each borehole, a piezometer and a 3-component geophone were installed. Additionally, nine 3component geophones were deployed in shallow 5 m holes distributed over the investigation area (Figure 2). With such a seismic network, accurate locations of microseismic events within the critical rock mass can be expected.

Figure 1: Overview of the Randa rockslide area. Black rectangle outlines investigation area shown in Figure 2.

Figure 2:
Air photograph of the investigation area. Brown square represents central recording site, yellow squares denote Geode seismographs and green square indicates position of data logger. Borehole positions are indicated with blue stars and geophone positions are marked with red dots. Analog seismic and data logger cables are schematically represented by red and green lines, whereas digital connections and power cables are shown by black lines.

Operating a monitoring system in a high-mountain area is a challenging task. Harsh weather conditions and difficult access impose a variety of logistical problems. Our setup is displayed schematically in Figure 3. Electrical power for the entire monitoring system is generated with solar panels and a wind generator (Figure 4). During the absence of sufficient sun radiation and/or wind, battery packs (24V/800 Ah) ensure autonomy for approximately 10 days

Figure 3: Schematic overview of the Randa monitoring system.

Data from the inclinometers and piezometers are collected by a data logger and transferred via a RS-232 interface to the central recording site, where an industrial field PC is installed. The twelve 3-C geophones are connected to two Geometrics Geode seismographs, which operate in event trigger mode. To minimize the lengths of the analog cables from the geophones to the seismographs, the seismographs are placed at appropriate locations within the network (Figure 2). The Geodes and the field PC communicate over a fast Ethernet connection. The field PC stores the data temporarily on a solid-state disk. Since more than 1.5 km of cables were deployed in the investigation area, special attention has to be paid to excess voltage protection (lightning storms). Figure 5 displays some technical details of the installation

Figure 4: Central recording site (brown square in Figure 2) with solar panels and wind generator.

The amount of data generated by the inclinometers and piezometers is relatively small, whereas a single event recorded with the seismic system requires 12 Mbytes of storage space. As a consequence, data are transferred on a regular basis via an approximately 2 km long wireless Ethernet connection to a server PC located in a building of the Randa local authorities in the valley. The server PC can be accessed by a telephone connection from ETH Zürich. This is suitable for supervising the system and for transfering small amounts of data. The large volumes of seismic data are dumped to tape and sent by mail to Zurich. A critical aspect of our data transfer scheme is the high level of redundancy. Data are stored (at least temporarily) on the field PC, the server PC and on tape. In case of a breakdown of either the wireless Ethernet connection or the server PC, or if the tape gets lost or damaged during transportation, continuity of the data stream can be guaranteed for a certain period. All of these possible incidences have occurred at least once during the past 16 months of operation, but no significant loss of data has resulted so far.

Figure 5: a) Geode installation (yellow square in Figure 2) with excess voltage protection. b) Field PC at the central recording site.

To date, approximately 400 Gbytes of seismic data have been collected. Most of the "events" have been identified as false triggers. Approximately 300 events may be related to activities within the rockslide mass. Data are currently being analyzed, such that first results can be expected soon. Depending on the outcome of the analysis, we plan to continue operation of the monitoring system for at least 2 years.

Acknowledgements
We are grateful to the local authorities of Randa for all their support. Furthermore, we owe thanks to J.D. Rouiller of CREALP (centre de recherches sur l'environment Alpin), who supported the project financially and logistically. The main financial contributions were provided by the Swiss National Science Foundation and ETH Zurich.

ETH Zürich research team
Hansruedi Maurer, Tom Spillmann, Beat Rinderknecht, Christoph Bärlocher, Heike Willenberg, Björn Heincke, Erik Eberhard, Keith Evans, Simon Löw and Alan G. Green.Corresponding author: Hansruedi Maurer (hansruedi.maurer@ethz.ch)

back to the start of the article "Expanded Capabilities for Geode Seismograph..............."

back to GeoProfiles newsmagazine.......................

[Home]