Riverbank Stabilization with Bioengineering: Failure Processes and Design Suggestions from a Scaled Physical Model

Soil and water bioengineering structures (SWBEs) implemented for riverbank stabilization are often designed based on field feedback and expert knowledge. These structures are weaker in the years just after implementation, before vegetation recovery, which limits their reliability and success rate. This study uses a 1:25-scale physical model to compare the failure processes of noncohesive riverbanks protected at their toe with riprap, early-stage fascines, or double fascines. The results highlight the key weak points that should be targeted to further improve SWBE design. In complement to qualitative observations, the topography is measured by photogrammetry before and after flooding with bankfull discharge. Pore pressure is measured in the channel bed and banks during the experiment. The results showed for banks with SWBE toe protection in meanders that (1) scour hole formation drives bank destabilization; (2) due to undermining (i.e., scouring), common SWBEs such as single-bundle fascines are insufficient to stop bank material falling into scour holes; (3) the bank toe can be stabilized by a combination of living structures (e.g., one bundle of fascines) and buried structures that play the role of filter, blocking bank material, and anchorage (here, a second bundle was used); (4) bank slope reduction reduces scour hole depth; and (5) in this scaled physical model without soil cohesion, bank erosion is driven by fluvial entrainment rather than by mass failure or water overpressure in the bank.