Fluid shear influences on the performance of hydraulic flocculation systems.

Gravity driven hydraulic flocculators that operate in the absence of reliable electric power are better suited to meet the water treatment needs of green communities, resource-poor communities, and developing countries than conventional mechanical flocculators. However, current understanding regarding the proper design and operation of hydraulic flocculation systems is insufficient. Of particular interest is the optimal fluid shear level needed to produce low turbidity water. A hydraulic tube flocculator was used to study how fluid shear levels affect the settling properties of a flocculated alum-kaolin suspension. A Flocculation Residual Turbidity Analyzer (FReTA) was used to quantitatively compare the sedimentation velocity distributions and the post-sedimentation residual turbidities of the flocculated suspensions to see how they were affected by varying fluid shear, G, and hydraulic residence time, θ, while holding collision potential, Gθ, constant. Results show that floc breakup occurred at all velocity gradients evaluated. High floc settling velocities were correlated with low residual turbidities, both of which were optimized at low fluid shear levels and long fluid residence times. This study shows that, for hydraulic flocculation systems under the conditions described in this paper, low turbidity water is produced when fluid shear is kept at a minimum. Use of the product Gθ for design of laminar flow tube flocculators is insufficient if residual turbidity is used as the metric for performance. At any Gθ within the range tested in this study, best performance is obtained when G is small and θ is long.

Method for quantitative analysis of flocculation performance.

The sedimentation rate and the post-sedimentation residual turbidity of flocculated suspensions are properties central to the design and operation of unit processes following flocculation in a water treatment plant. A method for comparing flocculation performance based on these two properties is described. The flocculation residual turbidity analyzer (FReTA) records the turbidity of flocculent suspensions undergoing quiescent settling. The fixed distance across which flocs must travel to clear the measurement volume allows sedimentation velocity distributions of the flocculent suspension to be calculated from the raw turbidity data. By fitting the transformed turbidity data with a modified gamma distribution, the mean and variance of sedimentation velocity can be obtained along with the residual turbidity after a period of settling. This new analysis method can be used to quantitatively compare how differences in flocculator operating conditions affect the sedimentation velocity distribution of flocs as well as the post-sedimentation residual turbidity.

Flocculation model and collision potential for reactors with flows characterized by high Peclet numbers.

A mechanistically-based model is developed to characterize flocculation in the context of flow regimes with high Peclet numbers such as would occur in serpentine flow reactors. These flow conditions are obtained in gravity-driven hydraulic flocculators without mechanical agitation that are an important component of sustainable water treatment systems where energy efficiency and cost are receiving increasing emphasis. The model incorporates a fractal description of flocs, estimates of floc separation distances, estimates of relative velocities of floc particles dependent on the relevant flow regime, and provides an overall prediction of the required reaction time for formation of settleable flocs based on influent turbidity, alum dose, and energy dissipation rate. Viscosity is shown to be significant for the early stage of flocculation and turbulent eddies are shown to be significant for the final stage of flocculation. The collision potential defined as the product of the hydraulic residence time (θ) and the cube root of the energy dissipation rate (ɛ), i.e., ɛ(1/3)θ, is shown to be a better predictor of flocculator performance than the conventional product of θ and the velocity gradient (G), i.e., Gθ.

The role of aluminum in slow sand filtration.

Engineering enhancement of slow sand filtration has been an enigma in large part because the mechanisms responsible for particle removal have not been well characterized. The presumed role of biological processes in the filter ripening process nearly precluded the possibility of enhancing filter performance since interventions to enhance biological activity would have required decreasing the quality of the influent water. In previous work, we documented that an acid soluble polymer controls filter performance. The new understanding that particle removal is controlled in large part by physical chemical mechanisms has expanded the possibilities of engineering slow sand filter performance. Herein, we explore the role of naturally occurring aluminum as a ripening agent for slow sand filters and the possibility of using a low dose of alum to improve filter performance or to ripen slow sand filters.

Enhancing slow sand filter performance with an acid-soluble seston extract.

An acid-soluble extract was obtained from Cayuga Lake (Ithaca, NY) seston and applied to slow sand filters at different application rates. Biological activity in the filters was inhibited with 3mM sodium azide. The filters were challenged with a synthetic raw water containing Escherichia coli. The Cayuga Lake seston extract (CLSE) fed filters removed up to 99.9999% of the influent coliforms while the control filter (no CLSE) removed 50%. Filter performance was correlated with the amount of CLSE applied to the filters.