SEDIFLOC(2)-PBM

Ab initio modelling of physicochemical and hydrodynamical effects on flocculation/deflocculation in biotic model systems


The behavior of many dispersions is strongly influenced by their flocculation behavior. It is obvious that a lot of interest exists in the understanding of the mechanisms leading to either the avoidance or the achievement of flocculation. In the final application of the proposed research programme, i.e. biological wastewater treatment, (de)flocculation is of central importance because floc formation allows to achieve separation of the biocatalysts from the purified wastewater in the cheapest way, i.e. through gravitation. Worldwide, the lack of insight in flocculation/deflocculation is recognised and the uncertainty with respect to these processes is considered to be one of the weakest links in current wastewater treatment practise.

Essentially, flocculation/deflocculation can be seen as a set of 2 counteracting conversions between large (loose) and small (compact) (aggregates of) particles. The conversion rates are dependent on a lot of (not completely understood) factors such as the chemical composition of the medium, the physical (shear) forces that are acting on the flocs, the particle concentration, etc. Under stable conditions, a dynamic equilibrium is reached leading to a distribution of floc properties like particle sizes. Fundamental studies aimed at predicting flocculation/deflocculation are mostly limited to a description of the size of the formed flocs under well-controlled circumstances and have little or no attention to porosity or the resistance to imposed shear forces (e.g. through hydrodynamic stress). As a result the predictive power remains limited.

The proposed research aims at obtaining fundamental insight into the relation between, on one hand, interparticulate interactions and, on the other hand, flocculation and floc properties. With such insight a model can be built that allows to predict floc properties within the practical context of wastewater treatment. The (de)flocculation is primarily determined by physico-chemical properties of the medium and the forces that act on the flocs. Both elements can be obtained via simulation of the hydrodynamics in the bioreactor. The integration of a submodel that describes the interaction between flocculation/deflocculation of dispersions within a hydrodynamical model in which the local shear stress and physico-chemical properties can be simulated, must allow to predict their interaction. To build this submodel, one has to gather detailed and quantitative information concerning the relation between floc properties on the one hand and particle properties (zeta-potential, floc size, concentration, porosity, fractal dimension) and medium properties (ionic strength, pH, pO2, shear stress,...) on the other hand. This knowledge is essential to simulate, and therefore to check, empirically established phenomena.

Computational fluid dynamics (CFD) give information (via Navier-Stokes equations) about the spatial pressure, velocity and temperature distribution in the reactor. Combined with turbulence models and elementary flux balances (incl. biological conversions) it provides perspectives towards the prediction of spatial variations of particles, physico-chemical properties and shear forces in the specified biorector. Such numerical models should include the mentioned submodel describing flocculation and deflocculation. Consequently, such integrated model must allow to predict the evolution of floc properties throughout the bioreactor as function of imposed disturbances (other physico-chemical properties or shear forces) and to investigate possible optimisations of the system under study.

Project duration: 2000-01-01 - 2003-12-31


Project home page

http://biomath.ugent.be/~sedifloc


Contact

Ingmar Nopens
Department of Applied Mathematics, Biometrics and Process Control
Coupure Links 653
9000 Gent
Belgium
Tel: +32(9) 264.59.35
Fax: +32(9) 264.62.20
Email: Ingmar.Nopens@biomath.ugent.be


Last update: 01 december 2008, webmaster@biomath.ugent.be


  zoek print log in nederlands