Failure mechanisms and pore water pressure conditions: analysis of a riverbank along the Arno River (Central Italy)

Mechanisms of failure occurring in two portions of a riverbank along the Arno River (Central Italy), are investigated in detail starting by a series of periodic field observations and bank profile measurements. Two dominant mechanisms involving the silty sand portion of the bank have been observed: (a) alcove-shaped failure in the middle portion of the bank; (b) slab failure involving the middle-upper bank. A portion of the riverbank was subject to laboratory (grain size analysis; phase relationship analysis; triaxial tests) and in situ tests (borehole shear tests (BSTs)) to characterise the geotechnical properties of the overbank deposits. Two different procedures of bank stability analysis have been performed: (1) a complete analysis, coupling seepage analysis with the limit equilibrium method; (2) two simplified analyses, through the limit equilibrium method with simple assumptions on pore water pressures distribution. For the complete analysis, saturated/unsaturated flow within the riverbank was modelled by finite element seepage analysis in transient conditions, using as boundary conditions eight hydrographs with increasing water stage. Riverbank stability analyses have been conducted by the Morgenstern-Price rigorous method, dividing each of the eight hydrographs in 21 time steps and calculating the safety factor for each step. The analysis revealed the occurrence of two possible mechanisms of failure (slab-type and alcove-shaped sliding failures), according to the field observations, related to different river stages and pore water pressures within the riverbank: alcove failures are likely to occur with moderate flow events, while slab failures are favoured by flow events with higher peak river stage. A first type of simplified analysis, representing critical conditions reached during a rapid flow event, was based on the main hypothesis of the occurrence of a zero pore water pressures zone within the portion of the bank between the low-water stage and the peak stage reached. A second type of simplified analysis was applied in order to represent rapid drawdown conditions following a prolonged flow event (worst case), with the main assumption of total saturation of the material up to the same elevation of the peak river stage. The first simplified analysis has given similar results to the complete seepage/stability analysis, confirming slab-type and alcove-shaped failure as the two main mechanisms of instability, while the second type of simplified analysis has conducted to too conservative results compared to the other previous analyses. Field observations regarding different characteristic bank geometries in adjacent sub-reaches have been summarised in a conceptual cyclic sketch, that include all the possible paths of bank evolution depending on the succession of river stages reached during flow events and related pore pressure conditions. (C) 2001 Elsevier Science B.V. All rights reserved.