Optimization of a Flocculation-Sedimentation Treatment Plant with Computational Fluid Dynamics

Transcription

1 Optimization of a Flocculation-Sedimentation Treatment Plant with Computational Fluid Dynamics Zweckverband Landeswasserversorgung Dipl.-Ing. Marian Brenda Hamburg

2 Backround Information The Zweckverband Landeswasserversorgung provides drinking water for the north-east of Baden-Württemberg and parts of Stuttgart Water delivery 90 Mio m³/a Within the water treatment plant of Langenau raw water from the river Danube is purified

3 Overview Capacity 400 to 800 l/s Diameter Accelator: 36 m Height Accelator: 9,5 m

4 The Sludge Model (Singlephase) Sludge is modeled with a user defined scalar (uds) within ANSYS FLUENT and contains of the following components: Convection Settling velocity depends on concentration Diffusion Diffusion depends on density gradients Viscosity Viscosity depends on concentration and shear rate Density Density depends on concentration

5 Measuring the Settling Velocity During the settling test the sludge level is measured over the time Within the first minutes linear settling takes place Thereafter hindered settling takes place Within the phase of linear settling the concentration and settling velocity is constant settling velocity for a certain concentration Relatively easy to measure (standard test DIN ) Height of the sludge level after 30 Minutes is called sludge volume (important parameter in waste water treatment plants) Sludge level (ml/l) linear settling Settling curve hindered settling Time (min)

6 Settling Velocity depending on Concentration v s v 1 0 h c n Settling velocity [m/s] 0,003 0,0025 0,002 0,0015 0,001 0,0005 Settling velocity depending on concentration This zone is difficult to measure (impossible with standard settling test) Concentratin [g/l] In fact the maximum settling velocity reaches ist maximum between 1 and 2 g/l. For lower concentrations the settling velocity decreases to 0 (not implemented in the presented simulation).

7 Diffusion and Turbulent Viscosity The diffusion of sludge is dominated by the turbulent diffusion, which is calculated from the turbulent viscosity: m t The turbulent viscosity is suppressed, when high density gradients occur (at the boundary layer between high concentrated sludge and clear water): G t g Pr t x i Therefore the exchange of momentum and mass (of sludge) tends to zero at the boundary layer stratified flow

8 Modelling the Settling Velocity and Diffusion n s c h v v 1 0 z z y c y x c x z v W y Vc x Uc t c s ) ( The settling velocity and Diffusion-coefficient are added to the transport equation: t m

9 Rheology Sludge is a non Newton fluid The viscosity increases with higher sludge concentrations The viscosity decreases with higher shear rates (shear thinning) Sludge shows time dependent viscous properties Difficulties with the measurement of the viscosity of sludge: Seperation (settling) of sludge during measurement especially when measuring sludge with lower concentrations Sludge is inhomogeneous and contains materials like sand particles, hairs, grease etc. Difficulties with modelling the viscosity of sludge: Modelling of the yield stress Thixotropic behaviour

10 Flow Curve Flow Curve Shear Stress [Pa] 0,07 0,06 0,05 0,04 0,03 0,02 0, ,4 g/l 14,8 g/l 12,6 g/l 10,8 g/l 7,2 g/l 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 Shear Rate [1/s] Rheological model of Worrall-Tuliani. Yield stress (τ y ) is set to 0 in the presented simulation. y D B y D D

11 Density The density depends on sludge concentration: w s w c s

12 Validation of Sludge Model The settling test was simulated with FLUENT The results were compared with the measurements Simulation and measurements fit quite well according to sludge volume Sludge Volume - Concentration Sludge Volume (ml/l) Concentration (g/l) Accelator Simulation

13 What the Sludge Model actually does

14 Flocculation-Sedimentation plant ( Accelator ) clear water stirrer inlet sedimentation zone flocculation zone

15 Optimization Several options were simulated to increase the capacity of the Accelator Deflectors to becalm the sedimentation zone Several configurations of lamella clarifier Unplunged lamellas showed the best results

16 Geometry Lamella clarifier (for testing purposes only in one section) Clear water chanel (outlet)

17 Velocity

18 Sludge concentration

19 Velocity Field between the Lamellas Clear water moves up Sludge moves down

20 Results Results can be seen only in comparison between the simulations. For validation on the real plant the lamella clarifier has to be built and measurements have to be taken to compare simulation with reality. Sludge concentration at the outlet of lamella clarifier are 2 (10cm distance between lamellas) to 5 (5cm distance) orders of magnitude lower than without Further research is necessary to improve the sludge model and to verify it with the real plant

21 Thank you for your attention Zweckverband Landeswasserversorgung

22 Flow regime in the sedimentation zone

23 Stratified Flow (Turbulent Viscosity) Suppressed turbulent viscosity at the boundary layer of clear water and sludge