Report No. : GEO Report No. 95
Report Title : The Lai Ping Road Landslide of 2 July 1997 (1999), 140 p. plus 2 drgs.
Author : H.W. Sun & S.D.G. Campbell
On 2 July 1997, a landslide occurred at a roadside cut slope (No. 7NE-C/C95) at Lai Ping Road, Kau To Shan, Sha Tin. The incident comprised several discrete failures along a 135m-long section of the cut slope and completely blocked Lai Ping Road. During the initial investigation of the landslide incident, a major scarp complex was found in the densely-vegetated natural terrain, some 30 to 50 m above the crest of the failed cut slope. Evidence of fresh ground displacement was observed, with estimated total vertical and horizontal displacements of the order of several metres. This feature defines the possible limit of a very large-scale landslide, the total volume of which is estimated to be about 100 000 m3.
Following the landslide in July 1997, the Geotechnical Engineering Office (GEO) carried out a comprehensive investigation to study the mechanism and causes of the failure. The study has established a history of instability of the slope, and has documented absolute movements of up to several metres at the ground surface. The likely phasing and magnitude of the movements were established by means of topographic surveys, conventional and digital photogrammetry analyses of terrain models constructed from aerial photographs, detailed field mapping of the main and minor scarps, ground investigation and dendrochronology.
The study has included detailed rainfall analyses, incorporating confidence limit assessments, to assess the return periods of the July 1997 rainstorm based on statistical extrapolation of past rainfall records. The assessed return periods were found to vary considerably, depending on the reference raingauge adopted. The uncertainties are substantial, with the assessed return periods falling between 40 years and 1 000 years.
The first reported major failure of the slope was probably triggered in June 1978 by a moderate rainstorm shortly after the cut slope was completed. Subsequently, the landslide retrogressed into the natural terrain upslope, resulting in stress relief and the development of tension cracks and scarps. These features, and natural soil pipes found in the western part of the landslide site, would have led to enhanced infiltration and subsurface groundwater flow. In the1997 failure, there is evidence both of renewed shallow failures at the face of the cut slope, as a consequence of infiltration, and retrogression of the large-scale landslide, triggered by a significant rise of groundwater level following the rainstorm on 2 July 1997.
The most probable failure mechanism for the large-scale landslide involves shearing close to the base of saprolite, coupled with a deep-seated zone of internal deformation resulting from localised shearing along discrete joints, as well as shearing through intact materials. The main cause of this instability is likely to have been the formation of the major cut slope in the steep natural terrain in 1978 which increased shear stresses in the weak and deeply weathered rock mass. An additional contributory factor is likely to have been stress relief in the ground mass generated by the initial and subsequent failures. The principal trigger of the landslide movements in different periods is likely to have been the build-up of pore water pressures due to elevation of the main groundwater table in a complex hydrogeological regime, following heavy rainfall.
The ground investigation, which included a number of deep drillholes, has provided no evidence of shear surfaces within the rock below the saprolite that could be related to landsliding.