ABSTRACT: Geotechnical hazard assessment of the south flank of Frank Slide, Hillcrest, Alberta.

CGRG Bibliography of Canadian Geomorphology
Search Results

Author :

Date :

Title :

Publication :

Issue :

Page(s) :

Abstract

The Frank Slide (McConnell & Brock 1904) on April 29, 1903, which involved 30 million cubic metres of the east face of Turtle Mountain, destroyed the southern end of the town of Frank, Alberta and claimed 70 lives. The slide lasted 100 seconds, and covered an area 3 km2 with an average of 14 m of rubble. The Frank Slide is a Provincial Historic Site within a Restricted Development Area. Studies following the 1903 Slide investigated its cause, and the likelihood of a similar event ocurring from the prominent North and South Peaks left after the removal by the event of the central peak. Factors contributing to the Slide have been identified as: · the geological structure of Turtle Mountain; · subsidence/caving due to coal mining at the toe of the mountain; · above-average precipitation in years prior to slide; · water and ice accumulation in cracks at the top of the mountain; · seismic activity in 1901 and blast-induced seismicity; and · thermal variations and freeze-thaw cycles. The 1903 event created a network of deep subvertical fissures at the top of Turtle Mountain. Monitoringof these fissures commenced in 1903 following the Slide. Between 1931 and 1933, J. Allan conducted three investigations of the stability of South Peak. A large and a small “danger zone” associated with failure of ~5 million cubic metres of rock from South Peak were identified. The study resulted inrelocation of residents from within the small “danger zone” to neighbouring communities. Subsequent monitoring of Turtle Mountain included crack aperture measurements (starting in 1933), displacement measurements using Moiré gauges, photogrammetry, EDM surveys and trilateral signs, and seismicmonitoring (Kostak & Cruden 1990). Regular monitoring was discontinued by the early 1990’s. In September 1999, the Alberta Energy and Utilities Board retained BGC Engineering Inc. to conduct a geotechnical hazard assessment of the south flank of the Frank Slide near Hillcrest, Alberta. The project was initiated by Alberta Environment and managed by the Alberta Geological Survey. This paper describes the field investigation and analyses of rockfall and rockslide travel conducted for the project. The field investigation conducted between September 21and 24, 1999 sought · to confirm the geological structure and fissure patterns (Allan 1933) in the South Peak area, · to assess the state of instrumentation previously installed on Turtle Mountain, and · to examine collapsed mine workings near the base of the mountain. The dominant geological features near South Peak are the Turtle Mountain Anticline, the Turtle Mountain Thrust, bedding, jointing, and the network of major fissures (Cruden & Krahn 1973). The South Peak isPaleozoic limestone of the Livingstone and Mount Head Formations (Rundle Group). The limestone is vuggy, with large karstic features along bedding on the west limb of the anticline. The Turtle Mountain anticline axis trends towards azimuth 006°, and is 100 m east of South Peak. The average axial planeorientation is 90°/096° (dip/dip direction). The 1903 Slide occurred along bedding on the east limb of the anticline, moving towards azimuth 060°. The Turtle Mountain Thrust an d splay form the toe of the 1903 rupture surface above the outcrop of coal associated with the Frank mine. This thrust is sub-horizontal where it intersects the ground surface, andwas an active element in the 1903 Slide. The thrust is present below South Peak, but the splay is absent. The major discontinuities in the vicinity of South Peak include bedding planes, flexural slip surfaces parallel to bedding, and three joint sets on the hinge of the Anticline orthogonal to bedding, plus other random joints. The dominant joint set has a dip direction of 060°, parallel to the existing cliff-forming set exposed on the east face of Turtle Mountain. According to Krahn & Morgenstern (1976), flexural slip surfaces and joints have friction angles ranging from 32° (peak) to 14° (residual). The fissures created by the 1903 Slide, some over 30 m in depth, follow joint sets and flexural slip surfaces. These have two dominant dip directions: 060° (parallel to the existing east face of TurtleMountain) and 100° (parallel to the anticlinal axial plane). These fissures are subvertical near surface, but dip towards the east face of Turtle Mountain at depth. Displacement on these fissures is in an easterly or north-easterly direction. These features have apertures of metres, and contain ice and standing water in spring. Remnant snow was observed in one of the major fissures in September 1999. During the field investigation, installed instruments and monitoring stations were located. Many of the instruments have been damaged or destroyed through vandalism. Six of Allan’s original 18 gaugingstations allowed crack aperture measurements. There were no observable differences in crack aperture from Allan’s 1933 measurements at these stations. Coal mining associated with the Frank Mine was carried out in the Kootenay Formation between 1901 and 1918. The Turtle Mountain thrust separates the older Rundle Group from the younger Fernie Group and Kootenay Formation. The coal seam exploited in the Kootenay Formation was approximately 4.5 m thick, oriented at 85°/270°. In mining by a room-and-pillar method, chambers were 40 m long and between 40 and 122 m high, and pillars were about 9 m wide. Pillar-robbing and other mining practices undertaken to raise coal production were criticized by mine inspectors. Mine workings below the thrust fault in the Kootenay Group show large surface subsidences. Further subsidence is anticipated as timber supports rot, the fire which closed the mine burns, and caving of the open rooms continues. The results from the field investigation were used to define a line of potential points of release of blocks for use in the rockfall analysis. The potential volume and associated rupture surface for use in the rockslide analysis were also determined. The rockfall analysis used a lumped mass 3D model based on the DEM. The model allows the definition of lines from which blocks are released, with specified ranges of initial velocity and initial angle at discrete points. The 3D trajectory of a roll-down is evaluated by calculating the impact forces before andafter each contact with the terrain, characterized by dynamic restitution coefficients. A criterion is built into the model to determine whether a roll-down bounces or slides and rolls after each impact. Roll-downs were simulated from between elevations 1860 and 2060 m on the east slope of South Peak.All the roll-downs tend to roll or slide rather than bounce. Bouncing occurs only on a short stretch of the overall trajectory of a few roll-downs. The roll-down distances vary between 140 and 1300 m from their origin. The roll-downs tend to concentrate at the end of the trajectories, thus confirming the existence of channels and chutes visible on air-photos. The extent of simulated rockfalls is shown in Figure 1. The sliding volume and rupture surface were determined by considering kinematic failure modes along different planes at different depths below South Peak. Based on results of the field investigation, two physically possible rupture surfaces were analysed. The choice between the two surfaces was made after a probabilistic failure development analysis (Oboni & Bourdeau 1983). The most likely sliding volume was estimated at 5 million cubic metres, resting on a rupture surface oriented at 32°/088° with curved upper and lower extremities.Rockslide travel was estimated using both 3D numerical modeling and empirical relations, and the DEM of the site. The numerical model used Cellular Automata to forecast the path of the movement and the final distribution of the flowing material. The single dry flow simulation represents a lower bound on thetravel of a rock volume of 5 million cubic metres. However, the results provide a good estimate of the range of possible flow directions and general shape of the accumulation. Empirical relations (Hutchinson 1988) provided a best-estimate (calibrated to the Frank Slide height/travel distance relationship) and upper and lower bounds (corresponding to the upper and lower limits of mobile sturzstroms defined byHutchinson 1988) of travel. The results of the analyses from the model and empirical relations are superimposed on a photo montage of the site, along with Allan’s original 1931 “danger zones” . The deposit obtained using the numerical model covers a surface of 1.5 km2 , 900 m wide at the toe. Elevation 1300 m is reached at the north end and 1350 m at the south end. The shadow obtained from the single rock roll-down simulation is similar in shape to that evaluated with the rock flow model, but does not extend as far. Transport distances yielded by Hutchinson’s empirical relation show, for the lower bound, an accumulation similar to that obtained with the model. The best-estimate of travel distance, based on observations of the Frank Slide, is about 250 m beyond the lower bound. The upper bound empirical estimate based on Hutchinson (1988) falls between the large and small “danger zones” defined by Allan (1931). However, Allan’s zones have a sharp promontory shape, whereas the empirical estimate describes a gentle arc 1800 m in length. Laterally, the extent on the slope is controlled by a topographic ridge to the south, and the original 1903 Slide path to the north. This study suggests that Allan’s 1931 estimates of “danger zones” are consistent in distal extent withmore modern empirical relations, but not in shape or lateral extent. Numerical modeling suggests that a completely dry flow will exhibit frictional behaviour similar to single blocks rolled down the mountain. However, empirical relations suggest that conditions different from those assumed for the dry flowsimulation, the inclusion of water or snow, and lower friction angles could result in greater travel. The maximum probable event, based on case histories of mobile sturzstroms, is defined by the outermost empirical boundary. Refinement of the lateral extent of such an event is possible through additionalmodel simulations.

Bibliography of Canadian Geomorphology