Modeling flocculation and deflocculation processes of cohesive sediments

The transport and fate of cohesive sediments are responsible for many engineering, environmental, economic and policy issues that relate to, for example, siltation and dredging in navigation channels, water quality, water turbidity, pollutant transports, and biological ecosystem responses. Our current understanding, however, is insufficient to conduct accurate quantitative predictions of these processes. This is because the cohesive particles in natural waters will flocculate, which determines the settling, and thus the deposition behaviors. The simulation of flocculation processes is a primary challenge since the time variation of Floc Size Distribution (FSD) is controlled by a partial differential equation that also contains the integration of FSD itself. Previous models either address less characteristic sizes, which produce biased FSDs, or are incapable of modeling a relative large study domain in order to better express the FSDs with more size groups. In this study, a cohesive sediment flocculation model developed based on the framework of Population Balance Model (PBM) is solved by the Quadrature Method of Moments (QMOM). This PBM�QMOM flocculation model has reasonably compromised by both the model robustness and model efficiency. The former lies in the capability of describing the time evolution of the FSDs with a maximum of eight size classes, and the latter is reflected in its efficiency to solve PBM with transport terms and the potential to be coupled in a flow-mud estuary model. The model predictions are compared to both the analytical (or trusted class method) results for general PBMs (i.e., beyond the scope of specific research field), and the published experimental results of kaolinite suspension and colloidal montmorillonite. After that, an experimental activity has been carried out to develop a Sony NEX-5R camera system (with extension tubes and close-up) to automatically acquire floc images under various controlled environments, and to use MATLAB software to process the FSDs. This process is validated by the results of two set of sample particles. The validated camera system is first applied in a five liter mixing chamber to investigate the effects of salinity and selected organic matters on kaolinite flocculation. Then, the camera system is improved and assembled in a waterproof house for underwater use to provide data for a conceptual one-dimensional application in a relatively large turbulence tank. The flow field of the tank is measured by an acoustic Doppler velocimetry. The flocculation processes in the mixing chamber or cylindrical tank are modeled by PBM�QMOM and validated by camera statistical FSDs. While chemical and biological effects are not explicitly included in PBM�QMOM (implicitly included in fitting parameters) at this time to address the basic mechanisms of flocculation, these effects can be further extended when the process itself is better understood through other laboratory experiments or field measurements