Experimental research on dynamic dredge overflow plumes

Abstract

The overflow mixture of hopper dredgers contains fine sediment which when released in the environment forms a so-called plume. The spreading of this sediment and the related turbidity increase and sedimentation pattern can be of environmental importance. Understanding the behaviour of these plumes is therefore important. The first phase of the plume is called dynamic. It is characterized by the rapid descent of a dense mixture under influence of its initial momentum and buoyancy. The dynamic phase is followed by the passive phase that is governed by the settling behaviour of the particles. The behaviour and fate of the dynamic phase is depending on several influences and is therefore more difficult to model than the passive phase. This M.Sc. thesis is focused on getting more insight in some of these influences. Also the implementation of these influences in dynamic plume models is investigated. A preliminary study was carried out to investigate both theory and modelling of dynamic plumes in order to define the knowledge gaps present. The possibilities to fill these knowledge gaps were studied which resulted in carrying out experiments on two interesting influences on the dynamic plume: stripping and vortex divergence. Stripping is defined as the removal of material from a dynamic plume by the cross flow. Its process is not well understood and the few quantifications presented in literature were found to have no proper base. To obtain more information laboratory experiments were performed at a 1:100 distorted Richardson scale. The experiments were partly hindered by the limited flume dimensions of the experimental setup, but the tests yielded interesting observations. Vortex divergence is depending on the strength of the vortex pair that is formed when a plume is bent over by a cross flow. In homogenous surroundings the plume is said to be non-bifurcating, but the vortices of the vortex pair are diverging at a near constant rate. This divergence was already observed in the stripping experiments and was further investigated in extra specialized experiments. Next to that the need and possibility for an extension of the CORMIX model to include this divergence was investigated.