Progressive geomorphic evolution of reservoir bank in coarse-grained soil in East China – Insights from long-term observations and physical model test

Highlights

• Bank collapse characteristics of a coarse-grained material reservoir were analyzed by conducting 5 years of monitoring.

• The three-dimensional evolution rule of bank collapse over time was revealed.

• A new parameter of stable angles, which can reliably predict BCWs, was proposed for coarse-grained soils.

Abstract

Reservoir bank collapse, as a typical geological hazard, severely affects the geomorphic evolution and safety of residential sites. In this study, a case of Tankeng Reservoir, which is a large reservoir in eastern China is used to analyze a three-dimensional evolution mechanism and deformation characteristics of reservoir bank collapse by years of field investigation, underwater topography measurements and monitoring and corresponding physical modeling experiment. The results show that 90% of the bank collapsed within six years, and a full collapse occurred within nine years for the coarse-grained soil bank. In addition, a three-dimensional progressive pattern for the bank collapse evolution process and a new parameter of stable angles were proposed for first time, and this parameter has practical significance for further scientific calculations of the width and length of bank collapse.

Introduction

With the large-scale development of hydropower construction in western China, immense changes have taken place in the water ecosystem (Tang et al., 2006a, Tang et al., 2006b; Fan et al., 2012; Xu et al., 2018; Ji et al., 2019). These changes have led to several natural disasters. Reservoir bank collapse, which is the gradual receding of the bank under the action of the reservoir water, is one of these hazards (Kachugin, 1949; Kapayxev, 1958). When the collapse width reaches a certain size, it threatens the safety of buildings and farmlands situated along the bank.

In terms of the prediction of the bank collapse width (BCW, refer to the horizontal distance between the trailing edge and the leading edge of the collapse area, its direction is perpendicular to the slope strike), Kachugin (1949) was the first to calculate BCW by using the underwater soil repose angle. Since then, the factors influencing bank collapse, such as soil compactness, slope angle, slope height, and wave and water level fluctuations, have been studied. In addition, many scholars revised the graphical model separately (Andrew et al., 2000; Mosselman et al., 2000; Wang et al., 2000; Couper and Maddock, 2001; Malik and Matyja, 2008). The construction of the Three Gorges Water Conservancy project in China has resulted in significant bank collapse within the reservoir. Because of an extremely large submerged area in the Three Gorges Water Conservancy project, the impact is very significant (Heming and Rees, 2000; Tan and Yao, 2006; Zhu et al., 2008). The construction caused the relocation of more than 1.2 million people. During the construction of China's hydropower station, a large amount of research was conducted. Xu et al. (2007) established a new prediction method for the BCW, called the slope structure method, and applied it to the Fengjie County of the Three Gorges Reservoir Area, considering the difference in soil structure. Chen et al. (2014) revised the critical height of wave niches and soil, based on the Kachugin method. Peng (2014) established a prediction method based on the balance principle of collapse and accumulation. These methods greatly advanced the research for bank collapse prediction and effectively guided the construction of the project.

With respect to the interaction between reservoir water and the bank, the mechanism of bank collapse has recieved much interest. The degrees of influence of geomorphology, soil properties, slope structure, and waves on bank collapse were revealed. A hydraulic model (Osman and Thorne, 1988; Throne and Osman, 1988; Darby and Thorne, 1996) regarding the effect of soil cohesion on the bank was established. Simon et al. (2009) simulated a basal erosion process caused by the river hydraulic force and riverbank destruction under the action of gravity. Nardi et al. (2013) established a relationship between the water flow and erosion resistance of the bank. Davis and Harden (2014) studied the characteristics of reservoirs that had experienced bank collapse. Wang (2003) showed that the collapsed bank block can delay bank retreat by scouring experiments, and a similar conclusion was drawn by Duan (2005) by conducting physical model experiments. Moran et al. (2013) studied the interaction between riverbank erosion and riverbed evolution in the Rhine river by conducting a series of physical model experiments. Ji et al., 2017, Ji et al., 2018 studied five key factors affecting the bank through a physical simulation and ranked these factors. These experiments effectively revealed the influence mechanisms of many factors on the bank collapse.

In recent years, new techniques, such as artificial neural and fuzzy inference system networks, have been employed to develop predictive models (Nardi and Rinaldi, 2010; Wang et al., 2010; Sahoo et al., 2011; Zhou et al., 2014; Wu et al., 2017; Xu et al., 2017), and researchers have conducted landslide susceptibility assessments by utilizing GIS approaches (Guo et al., 2020; Li et al., 2019). Wang et al. (2016) established a prediction method for the two-stage bank collapse program based on the Microsoft Visual Studio and Arc GIS Engine. Wu and Shao (2015) predicted bank collapse using the back propagation (BP) neural network. These new methods have greatly advanced the research and achieved good results.

In summary, extensive research has been conducted on the factors affecting, collapse mode, and engineering hazards. It is known that the bank collapse of coarse-grained soil weakened over time gradually and finally tends to stop. However, this understanding has been in the qualitative stage for a long time, which lacks the support of quantitative monitoring data for the specific evolution time. This study will answer this question with accurate data. In addition, the stable slope angle of gravel soil bank slope underwater is one of the key parameters that affect the accuracy of calculation results of BCW. However, the value of this underwater parameter is not accurate over a long period of time. In the study, the more accurate stable slope angle of gravel soil underwater will be measured through multiple measurements during the period of reservoir water level falling. In addition, a significant amount of research was carried out before a reservoir impounded water for storage, and less attention was paid to what happens after storage. Whether the results are actually consistent with earlier predictions needs to be verified through tracking investigation, which is lacking in the related research. In addition, the progressive collapse process after water storage is not clear. The main innovation of this study is to explain the relationship between bank collapse evolution and time after a reservoir impounds water for water storage.

In this study, the Tankeng Reservoir in eastern China was considered as an example to compare predicted bank behavior with actual erosion and geomorphic changes after impoundment. Based on the basic geological conditions of the study area, the relationship between the bank collapse characteristics and time was revealed through five years of tracking investigation. Then, a physical model experiment was conducted to reproduce the progressive collapse phenomenon of bank collapse and reveal the three-dimensional evolution mechanism. The results provided a reliable framework for the soil–water interaction and scientific prediction of the bank collapse range.