Alonso G Griborio, PhD, PE 1 Randal W Samstag, MS, PE, BCEE 2. Hazen and Sawyer, Hollywood, FL, US

Size: px

Start display at page:

Download "Alonso G Griborio, PhD, PE 1 Randal W Samstag, MS, PE, BCEE 2. Hazen and Sawyer, Hollywood, FL, US"

Gwendolyn Terry
6 years ago
Views:

1 Alonso G Griborio, PhD, PE 1 Randal W Samstag, MS, PE, BCEE 2 1 Hazen and Sawyer, Hollywood, FL, US 2 Civil and Sanitary Engineer, Bainbridge Island, WA, US Introduction Clarifier Modeling Options Role of Models Calibration and Verification Definitions 2-D and 3-D Modeling General Modeling Process Model Calibration 2-D Case Study 2Dc Model 3-D Case Study Square Clarifier Field Test and CFD Radial clarifier EDI Evaluation Case Histories 1

2 SOR ClarifierM odelingo ptions Box models SOR+UFR 1-D models 2-D models 3-D models Increasing rigor Xj Vs j+1 j j-1 Effective use of all levels of models UFR 1-D Model Level Strength Application Weakness Hazen S im ple Discretesettling(Grit) Ignoreshydrodynam ics L im it S tate S im ple Zonesettling (S S T prelim inary design andoperational) 1D Com putationspeed A ltypesofsettling including2 phaseflow s 2D 3D Com putationspeed com paredto3d;runson laptop. Com pletenessof governingequations; highspatialresolution. A lclarifiersw herethere isadom inantflow direction;dynam ic sim ulations. A lclarifiers.s teady state sim ulationsw herea dom inantflow direction cannotbeassum ed. Ignoreshydrodynam ics Ignoreshydrodynam ics Ignoreslateralnonuniform ity insolidsand m om entum. L ongexecutiontim es; highlevelofexpertise required. McCorquodale et al. WWTmod 2010 Presentation 2

3 Calibration: Initial trials to adjust model parameters to reproduce field conditions (either long term data or field testing data). Verification/Validation: Tests to confirm that a model is representing field conditions. For example by independent stress tests with different flow or settling conditions or operating data 3

5 General Purpose: Predict clarifier performance and capacity Based on: Influent parameters (Influent flow and MLSS) Operational parameters (Units in

4 Calibration depends on model configuration (sub-models) and data collection Special sampling/data collection may be required for model calibration/verification Wicklein et al. (2016) Good modelling practice in applying computational fluid dynamics for WWTP modelling WST 73.5 General Purpose: Predict clarifier performance and capacity Based on: Influent parameters (Influent flow and MLSS) Operational parameters (Units in service, RAS flow) Sludge characteristics (settling, flocculation) Clarifier Optimization Radius = 60 ft 8.5 ft 14.8 ft 14 ft Slope = 8.33% Concentration (g/l):

Inputand operation param eters: Flow measurements; SOR, RAS MLSS Historical data Plant records Special

$fractionation properties Historical data Special sampling; Stress testing Dye testing; velocity$ $Discrete settling/fractionation test Inputand O utput P aram eters for M odel Calibration and$ Verification Flocculation parameters Sludge Volume Index (SVI) Settling properties of the sludge:

Verification Flocculation parameters Sludge Volume Index (SVI) Settling properties of the sludge:

5 Inputand operation param eters: Flow measurements; SOR, RAS MLSS Historical data Plant records Special field sampling Stress testing P erform anceparam eters: Effluent TSS; Blanket depth Solids profile; Internal Hydrodynamics S ludgecharacteristics: Settling and compression properties Flocculation and fractionation properties Historical data Special sampling; Stress testing Dye testing; velocity measurements Sludge profiling Sludge volume index (SVI) Settling/compression test Flocculation test Discrete settling/fractionation test Inputand O utput P aram eters for M odel Calibration and Verification Flocculation parameters Sludge Volume Index (SVI) Settling properties of the sludge: zone, and compression rates Discrete settling parameters Flocculated suspended solids (FSS) and dispersed suspended solids (DSS) 5

dc dt 3 C C 1 1 2 k ds 1 2 2 1 2 d d 1 2 2 C d 1 2 V S 2 V S 1 Shear-induced

6 Vesilind Equation V s V o e kx Takac s equation V s V o e X X k X e k 1 min 2 X min McCorquodale s five-component model Differential settling flocculation: dc dt 3 C C k ds d d C d 1 2 V S 2 V S 1 Shear-induced flocculation dn dt K m B * X * G K A * X * n * G Floc Breakup Floc Aggregation 6

S 0 ( K X TSS ) Solids-liquid interface height (cm) 30 25 V 20 s 15 10 5 0 0 20 40 60 80 100 Time (min) ZSV (m/h) 8 6 Vs

(Field Data) 45 Settling Velocity (ft/h) 40 35 30 25 20 15 10 5 Vs= Voe (-kx ) Vs(ft/h)= 47.0 e (-0.

7 S 0 ( K X TSS ) Solids-liquid interface height (cm) V 20 s Time (min) ZSV (m/h) 8 6 Vs = 8.50e X R 2 = Concentration (g/l) Field Data Expon. (Field Data) 45 Settling Velocity (ft/h) Vs= Voe (-kx ) Vs(ft/h)= 47.0 e ( X (m g/l )) y = e x R² = Sludge Concentration (g/l) Daily Average - SVI 9/8/ ml/g 9/9/ ml/g 9/10/ ml/g Average 150 7

8 Dye Conc. A Plug Flow Cont. Flow Stirred Tank B Dye Conc. Dye Conc. Ta Ta Time Time Reactor Configurations and Flow Curves C Arbitrary Flow Ta Time Dye tests provide an estimate of field performance of secondary clarifiers Compare to an ideal plug flow reactor Evidence of short-circuiting and density currents 8

Collect a MLSS sample (About 15.0 Liters) Use a six-paddle stirrer and fill each jar with 2.0 L of mixed liquor (avoiding unnecessary delays) Assign a flocculation time to each jar, e.g., 0, 2.

9 Collect a MLSS sample (About 15.0 Liters) Use a six-paddle stirrer and fill each jar with 2.0 L of mixed liquor (avoiding unnecessary delays) Assign a flocculation time to each jar, e.g., 0, 2.5, 5, 10, 20, 30 minutes Mix the samples at a G of approximately 40 s -1 Allow the sample to settle for 30 minutes Take a supernatant sample from each jar Measure the TSS 70 Supernatant SS (mg/l) C = a+(c O -a) e -kg t X C = (60.7) e t K A = 7.4 x 10-5 L/g SS K B = 8.00 x 10-9 s X = 2,800 mg/l G = 40 s Flocculation Time (min) Equation 2.39 Data n t K B G K B G K A X G t no e K A K kgt X C a ( C A O a) e Wahlberg et al. (1994) La Motta et al. (2003) 9

10 If possible, stress testing provides information on how the clarifiers react at maximum flow conditions When does failure occur? Useful information for modeling 10

11 Developed at the University of New Orleans in 2004, 2Dc is a 2-D CFD model customized for clarifiers It account for all the major process occurring in settling tanks (e.g., hydrodynamics, flocculation, environmental impacts) It accounts for the dynamics of the sludge inventory It can predict effluent quality and RAS concentration Allows visualization of the internal conditions in the clarifier, like position of the sludge blanket and flow pattern It incorporate the geometry and other internal features of the clarifiers Radius = 60 ft 8.5 ft 14.8 ft 14 ft Slope = 8.33% Concentration (g/l): Radius = 60 ft 8.5 ft 14.8 ft 14 ft Slope = 8.33% Concentration (g/l):

useful life 2-D Case Study Field testing performed to understand How the SEP clarifiers respond

12 57 mgd ADWF HPOAS facility Design WW Flow = 250 mgd (150 mgd through secondary treatment) Sixteen (16) 120 ft Diameter Circular Center Feed Peripheral Overflow SCs SC reaching end of useful life 2-D Case Study Field testing performed to understand How the SEP clarifiers respond to stress conditions Determine MLSS settling, flocculation and settleability parameters Day 2 Day 3 12

13 2-D Case Study The following data were collected and analyzed: Stress Testing: Influent and effluent flows Return activated sludge flow rate MLSS, RAS SS, Effluent SS Sludge Blanket Sludge Volume Index Flocculation parameters Flocculated suspended solids (FSS) Dispersed suspended solids (DSS) Discrete settling parameters Settling properties of the sludge: zone, and compression rates Follow up testing Dye Studies RAS drawdown testing Stamford Baffle stress testing Microscopy 2-D Case Study Clouds of solids are an indication of poor clarifier hydrodynamics (strong internal currents) Cloudinessobservedw ithfourclarifiers online SOR 850 gpd/ft 2 ) S ignificantcloudinessobservedw iththreeclarifiers online SOR 1,000 gpd/ft 2 ) 26 13

14 11/11/ D Case Study SVI = 85 ml/g Effluent TSS = 29.4 mg/l RAS TSS = 13,450 mg/l 0.5 ft/s Center Well Diameter = 24-ft 720 Effluent TSS = 26.9 mg/l Conc. (mg/l) RAS TSS = 13,450 mg/l Min 0.5 ft/s Center Well Diameter = 30-ft A computational fluid dynamics (CFD) model was developed and calibrated and validated against stress testing data Model used to evaluate potential upgrades to be incorporated during clarifier rehab, including: inutes 720 Effluent T S S = 30.5 m g/l R AS T S S = 13,500 m g/l Min 0.5 ft/s Center Well Diameter = 36-ft Removal of inlet target baffles (were resulting in floc breakup) Center well diameter (existing too small) Stamford baffle position M 2-D Case Study Secondary Clarifier Capacity (MLSS = 2,800 mg/l, SVI = 85 ml/g, RAS = 30%) EffluentT S S (m g/l ) m g/l W eekly averageo utfalleffluentlim itation duringdry w eather Flow R ate-o neclarifiero utofs ervice(m GD) Existing Clarifiers Retrofitted Clarifiers

15 Samstag et al. (2010) 15

16 S ettlingvelocity T ests DyeT ests S olidsp rofilefield T est CFD S im ulationvelocity and S olids P rofile 16

Radial flow clarifier Questions: Optimum Depth?

1,100,000 hexahedral cells K-epsilon turbulence

17 Radial flow clarifier Questions: Optimum Depth? Optimum Inlet? Optimum Feedwell? Optimum Effluent Zone? Samstag and Wicklein (2014) 3D Fluent CFD 1,100,000 hexahedral cells K-epsilon turbulence model User defined functions (UDF) to implement Solids settling and transport Density coupling 17

18 11/11/2017 Field S olidsp rofilet est CFD offield T estconditions 18

19 11/11/

20 11/11/

21 11/11/2017 It is still necessary to perform calibration and verification tests for CFD evaluation of sedimentation. Calibration test examples: Settling tests Floc tests Verification test examples: Solids profile tests Dye tests CFD can uncover significant potential capacity and performance improvements 21

THINK FLUID DYNAMIX CFD Simulation of Clarifiers. THINK Fluid Dynamix

THINK FLUID DYNAMIX CFD Simulation of Clarifiers Provided by: THINK Fluid Dynamix Am Pestalozziring 21 D-91058 Erlangen (Germany) Tel. +49 (0)9131 69098-00 http://www.think-fd.com CFD ENGINEERING & CONSULTING