Study on reservoir slope

Since Tangyanguang landslide in China and Wayan reservoir landslide in Italy, the problem of reservoir landslide has been highly valued by the water conservancy and hydropower engineering circles and geotechnical engineering circles, and the research on this problem has been greatly strengthened in the process of reservoir site selection, feasibility demonstration and project operation. In large-scale water conservancy and hydropower projects, the revival of reservoir landslides, especially ancient landslides, is a common and very important problem, such as Lijiaxia Reservoir 1 No.2, Longyangxia Reservoir Longxi Landslide, Chadong Landslide, Baota Landslide, Baihuanping Landslide, Huanglashi Landslide and Jibazi Landslide in the Three Gorges Reservoir Area. In the process of reservoir impoundment, whether these landslides are resurrected, the scope of resurrection, the movement mode and dynamic process after resurrection are undoubtedly worthy of attention, and are also the basis of surge prediction.

Reservoir landslides can be divided into two types: one is the landslide under the influence of water and rock in the reservoir area; The other is natural landslide, which has no direct effect of reservoir water, but sometimes slides or collapses due to human engineering activities in reservoir resettlement development zone. The main mechanism of reservoir landslide is the hydraulic softening of sliding surface and the decrease of effective stress caused by the increase of seepage pressure on sliding surface. The most important factors of reservoir landslide are softening, mudding, undercurrent, dynamic water and hydrostatic pressure. There are two kinds of reservoir landslides: one is the landslide when the pore water pressure distribution in the landslide reaches a new balance after the reservoir water level reaches a sensitive level; The other is when the reservoir water level drops, especially in the period of rapid decline. The influence of reservoir water level fluctuation on slope stability is mainly related to the permeability and porosity of rock and soil. The rapid decline of reservoir water level causes instability in two aspects: one is unfavorable groundwater gradient; The other is excess pore water pressure. In addition, the erosion of the lower part of the landslide by the reservoir water and the internal corrosion and piping caused by the fluctuation of the reservoir water level have played a positive role in the stability of the reservoir bank and the resurrection of the landslide.

Because water is one of the most active and positive factors in the process of landslide, the research on hydrological function of landslide has been highly valued in the international landslide academic circles. The direction of groundwater seepage plays an important role in slope stability. Rainstorm infiltration leads to the decrease of matrix suction and soil strength, which leads to the occurrence of shallow landslides. Groundwater flow will form piping in sloping soil, which will reduce the cohesiveness of soil and lead to landslides. Although there are different views on the function and manifestation of water in landslide, generally speaking, the landslide caused by hydrological action is mainly due to the hydraulic softening of sliding surface and the decrease of effective stress caused by the increase of seepage pressure on sliding surface. During the rainstorm, the infiltration of rainwater leads to the increase of soil saturation, the sharp drop of suction and the sharp drop of shear strength. Persistent heavy rain will also lead to the rise of groundwater level or a short-term groundwater flow above the relative aquiclude. Therefore, when the duration and intensity of continuous rainfall exceed a certain limit, it will cause slope instability or landslide revival. Slope with a certain geological background and mechanical environment needs a certain amount of rainfall, rainfall intensity or rainfall duration to lead to slope failure. Therefore, at present, the forecast of rainstorm landslide is mainly based on the historical rainstorm-landslide records, establishing the empirical statistical relationship of critical rainfall intensity-duration combination, and monitoring and forecasting rainfall according to this relationship, so as to make the time forecast of regional landslide. However, this method often needs to record landslides and rainfall for a long time. However, in the same area, the combination relationship between rainfall intensity and duration triggering landslides with different material components and scales is often different, so it is often a macro criterion and cannot be applied to specific slopes or landslides. In order to know under what rainfall conditions a particular slope may be destroyed, or a particular landslide may be revived or partially revived, not only the physical and mechanical properties of slope soil, especially the potential sliding surface, should be systematically tested, but also the changes of pore water pressure of slope soil should be quantitatively described during rainfall. According to the physical mechanism of landslide induced by rainstorm, some scholars put forward leaky bucket model and flume model. Although these models are suitable for a single slope or landslide, the empirical parameters (such as drainage coefficient) in the models greatly limit their practicability. It should be pointed out that the above research on the role of water in landslide is generally limited to rainstorm landslide, and the research on the mechanism and quantitative process of reservoir water level change on landslide is extremely rare.

In order to predict the occurrence time of a specific single landslide, the field monitoring method is often used. On this basis, the occurrence time of landslide is predicted according to the macroscopic phenomenon of the surface and the changing trend of displacement-time curve of landslide deformation precursors, and some achievements have been made, such as the successful prediction of Xintan landslide in the Three Gorges of the Yangtze River, Jimingsi landslide in Zigui, Hubei, and Huangci landslide in Yongjing, Gansu. However, this method is only suitable for the prediction of impending landslide, and it is difficult to give the hydrological conditions of possible landslide revival quantitatively or semi-quantitatively. At the same time, due to the lack of systematic and comprehensive monitoring of most landslides, it is generally limited to the deformation monitoring on the surface or inside of the landslide. In the initial stage of landslide deformation, the monitoring frequency is usually once a month or even once a few months, so it is difficult to use the obtained data to deeply explore the hydrological mechanism of landslides.

In China, the study of landslide in the Three Gorges reservoir area is the most in-depth During the Sixth Five-Year Plan and the Seventh Five-Year Plan, a lot of exploration and research were carried out, and the stability of some large landslides before and after the dam was completed and impounded was analyzed and evaluated. It is considered that the relatively thick loose accumulation landslide has poor stability and deformation, and may be seriously deformed after impoundment. As for the monitoring and treatment of major landslides, since 1993, the treatment projects of Lianziya dangerous rock mass, Huanglashi and other landslides have been implemented, and the monitoring of Huanglashi landslide has also started since 1995. In addition, the counties and cities in the reservoir area have established landslide and debris flow monitoring workstations or geological workstations. However, due to the complexity of the problem, the research results, especially the prediction after the change of hydrodynamic environment, are still qualitative or semi-quantitative, and the research on the possible resurrection range of landslide, whether the movement mode after resurrection is slow crawling or fast sliding and its dynamic process is not deep enough.