Regular maintainance dredging is a large expense to the Port of Rotterdam, therefore the installation of sediment traps is considered. By increasing the bathymetry locally, an increase of local accumulation is expected and a decrease deeper in the harbour basin. This thesis describes the functioning of sediment traps in a stratified tidally energetic estuary, where sediment is supplied by the river and sea. A numerical 2DV representation is set up for the Botlek Harbour with the hydrostatic Delft3D software. A calculated salinity time-series by Operationeel Stromingsmodel Rotterdam (OSR) is combined with a measured water level time series for the same time period to describe the hydrodynamic boundary conditions. Simulations are done with a variable and constant Suspended Particulate Matter (SPM) time series boundary conditions, that were generated based on the measurements of De Nijs (2012). Flow expansion caused by the sediment trap reduces the flow velocity, but increases the turbulent kinetic energy locally. A reduction of bed shear stress is observed in the trap, except near the edges where an increase is observed. Density currents caused by salinity differences govern the vertical flow velocity distributions, while tidal filling causes the net exchange. The depth of the trap plays a significant role in the internal flow characteristics. The sediment trap changes the properties of the internal flow, which may change the hydraulic state of the flow, i.e. from supercritical to subcritical, resulting in large instabilities and even an internal hydraulic jump. Shallow sediment traps result in less frequent internal hydraulic jump and weaker jumps compared to deeper sediment traps. To investigate the dominant mechanism for the trapping of sediment in the sediment trap, a distinction is made between an erosion and a fluid mud scenario. The erosion scenario shows the largest agreement with the survey data, i.e. maintenance dredging data and Echosounder multibeam surveys, but contribute only marginally to the trapping of sediment. The fluid mud scenario yields the largest contribution to the trapping of sediment. The decreased amounts of accumulated sediment in the basin for substantial to lots of fluid mud behaviour may vary between 10% and 14%, depending on how this fluid mud behaviour is modelled. Various shapes have been tested on erosion and fluid mud scenarios. Both for erosion and fluid mud a more extreme choice of parameters might yield also more extreme results, this is however not considered realistic. The creation of an overdepth has shown to result in a marginal improvement in the capturing of sediment in an environment where erosion is important. A regular sediment trap decreases accumulation in the harbour basins by 2%. A trap twice as shallow increases this amount to 4%, but deepening the trap further may even enhance accumulation in the basins. A shorter or longer trap did not improve the situation. Installation of a sill did however result in a decrease of 6% compared to the situation without trap. It is concluded that this effect can be largely contributed to the internal flow properties and the presence of internal hydraulic jumps. The presence of an overdepth results in a significant improvement in the capturing of fluid mud flow. For the trapping of fluid mud flow, an overdepth results in 6% less sediment in the harbour basins no matter the depth or shape of the trap. The length of the trap did influence the accumulation in the basins. A trap twice as short shows an increase of 4 % of accumulation in the basins compared to a regular trap. A trap twice as long or installation of a sill results in similar accumulation in the harbour basins as a regular trap. For fluid mud flows, any type of overdepth decreases accumulation of sediment considerably. The amount of overdepth or the shape does not influence this. The length of the trap should be sufficient. A trap that is too short results in an increase of accumulation in the basins. For environments where erosion is important, only shallow sediment traps have proven to reduce accumulation in the basins. Traps that are too deep actually increase accumulation in the basins. A sill has proven to be the best measure for both mechanisms. If overdepth is desirable for navigation a shallow sediment trap is advised. This leads to a reduction of accumulation in the harbour basins for both erosion and fluid mud scenarios.

Phenomena like sea level rise, global warming and erosion together contribute to increasing flood risk in vulnerable coastal areas. As this flood risk increases, initiatives to mitigate the effects of climate change in the coastal zone are also increasingly sought. During recent years, the view that these measures should be circular has gained significant momentum. In line with this, the idea that sediment should not be treated as waste, but as a valuable product in a circular economy has been widely accepted. Within this theme, the sediment uses as resources in circular and in territorial economies (SURICATES) was created to develop and execute eco-innovative solutions for the reuse of sediments in Western Europe. As a part of SURICATES, over the course of a nine week pilot, a total of 500 tonnes of sediment was reallocated in a designated area in the Rotterdam Waterway, with the expectation that this sediment would be transported out of the Rhine-Meuse estuary into the North Sea. The SURICATES project is one of the first real large-scale efforts to reuse sediment for economic as well as ecosystem services. Effective implementation of the SURICATES project, however, requires a thorough understanding of the governing physical processes impacting the distribution of the deposited sediment locally. Furthermore, to assess the feasibility of similar future applications, efficient modelling of the pilot project is necessary. The work presented here aims to elucidate the hydrodynamic processes that are governing in the Rotterdam Waterway, within the framework of the SURICATES pilot project, and assess the reproducibility of these processes by (two) predictive models that are currently operational at the Port of Rotterdam. A special 6-hour monitoring survey was set up to measure salinities, velocities, temperature, and suspended particulate matter (SPM) along a transect crossing the reallocation location. In combination with a literature review, this dataset provides the basis for research into the predictive capabilities of two currently operational hydrodynamic models. Analysis of this dataset reveals the dominant terms in the momentum balance, the influence of Coriolis, the occurrence of internal waves, and the effect that all these mechanisms may have on the SPM distribution around the reallocation location. When the system dynamics are elucidated, the model performance of the two hydrodynamic models is assessed—both quantitatively and qualitatively. It is also investigated whether phase shifts are introduced in the models. It is found that the primary hydrodynamic processes in the Rotterdam Waterway are related to the barotropic tidal asymmetry imposed at the river mouth, the tidal excursion of the salt wedge, baroclinic exchange flow processes, and turbulence damping at the pycnocline. Turbulence damping at the pycnocline generally poses an upper limit to the (re)distribution of SPM over the water column, although field data suggests that this damping may not be sufficient to counteract diffusion processes locally. This effect occurs under certain forcing conditions and during low water slack. Furthermore, an internal Froude number analysis provides evidence for the possible generation of internal waves in one of the river’s bends. The evaluated models, however, are not capable of reproducing all of these hydrodynamic processes adequately. Although both models adequately reproduce water levels and the vertical velocity structure, they have difficulties predicting the pycnocline height. Additionally, it is found that both models introduce a small phase shift in the velocity and salinity prediction. The research presented here is a contribution to the understanding of the governing hydrodynamic processes in the Rotterdam Waterway, and the effect that (in)accurate modelling of these processes may have on future studies. Recommendations following from this research could improve future modelling practices.

The port of Rotterdam is located within the Rhine-Meuse estuary where a substantial amount of fine sediment transport takes place. Therefore, the port of Rotterdam is subject to significant siltation, requiring maintenance dredging to guarantee a sufficient nautical depth of fairways and harbour basins. To optimise the dredging strategy in the port of Rotterdam, a pilot study has been carried out wherein sediment is reallocated in the Rotterdam Waterway, during ebb, instead of offshore in the North Sea. Between May and November 2019, 210,000 tons of sediment has been reallocated. This pilot study has been carried out in the context of the larger EU-Interreg Sediment Uses as Resources in Circular and Territorial Economies (SURICATES) project. The main goals are to re-use the sediment as a resource and to reduce the sailing time of the dredging vessels. Both ideas comply with the Building with Nature philosophy; a concept gaining popularity over recent years in The Netherlands focusing on, amongst others the optimisation of dredging strategies. It is expected that the reallocated sediment is mainly transported offshore, while at the same time some of the sediment will nourish the river banks of the Rotterdam Waterway enhancing its flood resilience. This thesis focuses on understanding the fine sediment behaviour of the SURICATES pilot project on two different scales. This is done by analysing measurements and model hindcasts. The measurement campaign is set-up by Deltares and the Port of Rotterdam. For the model hindcasts an operational hydrodynamic and sediment model is used. On the small scale this is done by focusing on the behaviour of a single disposal over a tidal cycle. The large scale focuses on the cumulative long term behaviour of all sediment reallocations. For the small scale two measurement surveys are analysed. In both surveys it is found that a sediment reallocation executed by bow coupling is subject to mixing up to halfway the water column. Subsequently, the sediment plume is advected around and below the pycnocline. Further measurements in the mid field are lacking, but it is hypothesised that the majority of the sediment settles during subsequent low water slack. For the other execution method, drawing the bottom doors, which is used to reallocate the majority of the sediment, useful measurements are absent. It is hypothesised that the majority of this reallocated sediment is confined in the salt wedge and therefore mainly transported upstream over time. To assess the long term behaviour of the cumulative behaviour of all sediment reallocations, a different measurement campaign is set-up. In this measurement campaign, bed samples are collected prior to and during the pilot study to determine the change of the bed composition. In this campaign an indication for increased sedimentation related to the pilot study is found for nearly all the sample locations. The short term model study is set-up to enhance the understanding of the short term behaviour of a sediment plume, to derive an accurate source term for the sediment disposals and to carry out a sensitivity analysis. This sensitivity analysis is executed to derive the influence of differences in disposal method, timing of disposal, and uncertainties in the model. It is found that the execution method has the largest influence on the critical sediment fluxes on the short term, followed by the timing of disposal. From the long term model hindcast, in which the entire pilot study is hindcasted, it is found that 27% of the total amount of reallocated sediment flows downstream from the location of disposal and 73% upstream. These estimations are in line with the hypothesis and long term measurement results, but a thorough calibration of the results is lacking. To conclude, in this thesis a pilot study utilising a different sediment reallocation strategy in the port of Rotterdam has been investigated. This study shows that majority of the sediment disposed, in the current set-up of the pilot study, is estimated to flow upstream. In the sensitivity analysis, it is predicted that this might be caused by the timing of disposal or method of execution. It is also found that the initial behaviour of the sediment plume and the long term measurement contain a large amount of uncertainties. As most important recommendation for future work an expansion of the current measurement survey is proposed with at least two fixed locations: one downstream and one upstream of the location of disposal. In this way sediment fluxes can be established, which can also be used to verify and calibrate the sediment model.