Home >
Documents >
Numerical Simulation of the Flow Pattern in the Aeration ... JAMIL GHADERI.pdf · Numerical...

Share this document with a friend

12

Transcript

ISSN No. (Print): 0975-1130ISSN No. (Online): 2249-3239

Numerical Simulation of the Flow Pattern in the Aeration Tank ofSewage Treatment System by the Activated Sludge Process Using

Fluent ProgramHazhar Hadad* and Jamil Ghaderi**

*Department of Civil Engineering, Mahabad Branch, Islamic Azad University, Mahabad, IRAN**Department of Civil Engineering, Mahabad Branch, Islamic Azad University, Mahabad, IRAN

(Corresponding author: Jamil Ghaderi)(Received 07 January, 2015, Accepted 13 February, 2015)

(Published by Research Trend, Website: www.researchtrend.net)

ABSTRACT: One of the most important parts of the process in water or wastewater treatment plant is in theactivated sludge system of the aeration unit. The analysis of the flow pattern and velocity distribution in thetank is one of the important parameters in increasing the treatment efficiency of the unit, thereforeidentifying the static locations and the places with high velocity can be an accurate design index of this unit.In the present study the simulation of an aerated wastewater treatment tank in terms of two-phase conditionshas been evaluated for the first time. One of the three tanks is the wastewater treatment tank located inPardis Nazlu through the networked Gambit program then the networking was measured by the fluentprogram. Since the water in the tank is still the simulation and velocity distribution is performed by the staticflow and through aeration with a pipe with 280.2 m/h velocity. The simulation was performed at the momentof the air conditioning; the rate and the change of pressure demonstrated an insufficient aeration pipes in thesystem and the mixing and refining operations are not done in the best manner. The ratio of aeration to thetank area is 1 to 574000. However, if this ratio was 1 to 100, the mixing would be performed properly. It isrecommended that the ratio of aeration to the tank area would be increased in the wastewater treatmentsystem.

Keywords: Flow pattern, Tank aeration, Activated sludge, Fluent

INTRODUCTION

All the communities in daily activities produce wastesubstances that may be solid, liquid or gas form. Liquidwaste materials are so-called sewers or sewage. Seweris the water that is contaminated by variousapplications. About 99.9 percentage of sewers is waterand only about 0.1 percentage of it is consist ofimpurities such as suspended solids substance andcolloidal and solution [1].Wastewater is not just sewage. All the water used in thehome that goes down the drains or into the sewagecollection system is wastewater. This includes waterfrom baths, showers, sinks, dishwashers, washingmachines and toilets. Small businesses and industriesoften contribute large amounts of wastewater to sewagecollection systems; others operate their own wastewatertreatment systems. In combined municipal sewagesystems, water from storm drains is also added to themunicipal wastewater stream.

Wastewater is about 99 percent water by weight and isgenerally referred to as influent as it enters thewastewater treatment facility. “Domestic wastewater” iswastewater that comes primarily from individuals, anddoes not generally include industrial or agriculturalwastewater [8].Understanding specifications of wastewater is of thebasic processes in designing sewage purificationsystems. Without reliable data in this case, designs willbe met with many problems such as increasing cost, andoperation and maintenance problems and loss inreaching purification goals [2].According to Miller's comments, any change incharacteristics of air, soil, water and food that haveadverse effects on health, environment, humanactivities and other organisms is called pollution andany water that loses its quality for particular use, isturned into sewers. So the sewer is water whichconsumed and during this process, some suspended anddissolved substances entered in it [3].

Biological Forum – An International Journal 7(1): 382-393(2015)

Hadad and Ghaderi 383

Sludge can be defined as mud or soft sediment resultingfrom settling of solid substances exists in sewage. Also,the sludge can be assumed as mixtures of massedgranular particles, which hydrodynamically exhibitsuch as a single particle, and observed that these massescan be present, have no contact with other masses [5].Growing urbanization, population growth and urbanand industrial development, problem as water shortage(water scarcity) occurs. To solve this problem,strategies have been contemplated for purifying existingwater in environment and transform it into safe anddelicious water before delivering industrial or urbandrinking water networks and purifying domesticwastewater and industrial effluents before drainingthem into the nature. The major results of raw watertreatment and waste water treatment are includingproducing water and treatment wastewater, sludge andbyproducts [7].The purified water for urban or industrial and

wastewater of treatment plant is transferred to theacceptor water or lands and sludge prior to finaldisposal is refined for re-use purposes. In conventionalwastewater treatment methods such as active sludgeprocesses and stalactite filter, a high volume of initialsludge is added into secondary sediment sludge (activesludge). In active sludge process, secondary sludge is amicrobial biomass which is produced by themetabolism of organic matter, and microbial product insediment wastewater is about 50% and about 20% ofthis mass returned to the system and with initial sludgeguided to disposal system. In Stalactite filters, thehydraulic load is less and less sludge is produced andthere is no ancillary returning recycle system.Generally, high volumes of sludge with a concentrationof 1 to 4 percent of solid substance are produced intreatment processes which are created one of the mainproblems of excretion. The reason is that the excesswaste sludge is a combination of organic matter andmicrobial cells that may be degradable with othermicroorganisms [6].Planning and designing of wastewater treatmentfacilities is first and effective step in theimplementation of efficient Waste water treatmentplants technically and economically. Obviously usingsystematic and efficient methods have a great impact onsuccess of the project. Waste water treatment plants,consisting of a chain of physical, chemical andbiological processes which The main purpose ofguiding them is the removal of contaminants presentfrom one side and remove the remaining organicmaterial in the sludge treatment processes on the otherhand, before unloading them into the environment. Onphysical filtration (mechanical) stage of wastewater, the

very coarse aggregate solid particles floating orsuspended existing in wastewater are removed firstly.Also isolation of sand particles, are often mineral andcontaminated with fats and oils, is done. The final stageof physical purifying is related to initial settling(sedimentation) pool. In some cases, to improveefficiency of the solid-liquid phase separation, a smallportion of the surplus returned sludge is returned toinitial sedimentation pool or Chemical coagulantsubstances are injected to wastewater at this point. Theability of loading organic pollutant substances dependheavily on the amount of sludge in the aeration pool,because the amount of sludge available at aeration pooldepends heavily on the performance of secondarysedimentation basin (pool) under varying hydraulicloading conditions and returned sludge [4].As Computational Fluid Dynamics (CFD) findsapplication in an increasing number of scientific andengineering disciplines, the validation of the software isoften assumed by the user. However, now thatcommercial CFD software has the capacity to solve avast range of engineering problems, the validationmaterial that accompanies them can only ever apply toa subset of these applications. While academicresearchers may well accompany CFD simulations withexperimental campaigns in order to test a new model orsolution technique, industrial users often view CFD asjust another design tool. The latter are required toassume that the software, if used correctly, producesresults that can be relied on [9].Wahl et al. (2000) compared the results of a CFDsimulation against those from a previously validatedpiece of software called Win Flume. They foundreasonably good agreement between the CFD and themuch simpler analytical approach of Win Flume [10].Faure et al. (2004) used a version of a commercial CFDcode that was modified in such a way that the upstreamand downstream boundary conditions were appropriatefor use in modeling rivers and canals [11].Sarkar and Rhodes (2004). They compared the positionof the free surface profile over a weir in a flume withthat predicted by a CFD simulation. For the single flowrate they considered, the agreement was good but theyasserted that further work was required at higher flowrates [12].Chen et al. (2002) used CFD modeling to model theturbulent free surface flow over a stepped spillway. Theresearchers’ results generally compared well withexperimental data, the position of the free surface waspredicted very accurately [13]. Tokyay (2005)investigated the effect of channel bed corrugations on ahydraulic jump experimentally [14].

Hadad and Ghaderi 384

Esmaili et al. (2010) compared the values of pressureand velocity measured in laboratory with valuessimulated by the fluent model. The results showed thatthere exists good agreement between the on-the weirflow patterns measured in laboratory and simulated bythe model [15].Karim et al. (2009) simulated wave transformation inporous structures to calculate the hydraulic performanceof a vertical porous structure. The model was developedusing the VOF and two-phase flow [16].Taylor (1944) first addressed the topic of open-channeljunction flow by focusing on the depth ratio betweenthe upstream branches and the downstream channel.The results are a momentum analysis that yields apredictive equation for the depth ratio. Taylor’s paper isimportant for its identification of the need of theoreticaldescription of the open-channel junction and thegroundwork it formed for future investigations [17].Weber et al. (2001) performed an extensiveexperimental stud of combining flows in 90º openchannel for the purpose of providing a very broad dataset comprising three velocity components, turbulencestresses, and water surface mappings [18].Huang et al. (2002) provided a comprehensivenumerical study of combining flows in open channeljunction and investigate the effect of the junction angleon the flow characteristics [19].A numerical model using finite-element methods waspresented by Chong (2006), analysis the open channeljunction and a simple comparison with experimentaldata for velocity profiles were done [20].Pirzadeh and Shamloo (2007) provide application ofFLUENT -2D&3D software in simulation of lateralintake flows. Comparisons have been made betweennumerical results and measured experimental velocitiesfor a lateral intake [21].Namaee et al., (2014) simulated the 3 dimensionalflows on a broad-crested side spillway in a numericalway. RNG K-ε model was used to simulate turbulentand fluid volume model to find free water surface. Thesimulation results were evaluated using experimentaldata and showed that existing numerical methods usingRANS are useful in designing side weirs [22].Ramamurthy et al. (2009) applied a − model todetermine various flow characteristics such as velocitydistribution, water surface profile and pressuredistribution over a sharp-crested weir in a rectangularopen channel [23].

Research purposes are:(i) Determining the path length of air mixing with thewater treatment tanks.(ii) Providing the ability of the Fluent program insimulation of the two-phase flow.(iii) The general study of flow patterns in aerationtanks.

MATERIALS AND METHODS

In this paper, first of all, the location of the free surfaceof liquid metal was tracked by solving the volume offluid-function equation, determining computationaldomain of current time. Second, flow field andtemperature field were simulated by using SMPLEalgorithm; and temperature was got by solving energyequation. Third, add a time step. Don’t stop repeatingthis process until the filling process finish. Specificprocess is illustrated in Fig. 1.FLUENT is the CFD solver for choice for complexflow ranging from incompressible (transonic) to highlycompressible (supersonic and hypersonic) flows.Providing multiple choices of solver option, combinedwith a convergence-enhancing multi-grid method,FLUENT delivers optimum solution efficiency andaccuracy for a wide range of speed regimes. The wealthof physical models in FLUENT allows you toaccurately predict laminar and turbulent flows, variousmodes of heat transfer, chemical reactions, multiphaseflows, and other phenomena with complete meshflexibility and solution- based mesh adoption [24].FLUENT solves governing equations sequentially usingthe control volume method. The governing equationsare integrated over each control volume to constructdiscrete algebraic equations for dependent variables.These discrete equations are linearized using an implicitmethod. As the governing equations are nonlinear andcoupled, iterations are needed to achieve a convergedsolution.Conservative form of the Navier-stokes equations usingthe finite volume method on structured orthogonal,Cartesian coordinate’s grid system. Turbulent flows canbe simulated in FLUENT using the standard k-ε, largeeddy simulation (LES), renormalization group (RNG),or the Reynolds-stress (RSM) closure schemes. Themodel optimizes computational efficiency by allowingthe user to choose between various spatial (Second-order upwind, third-order, QUICK) discretizationscheme [25].

Hadad and Ghaderi 385

Fig. 1. Flow of Calculation.

The presence of turbulence is determined by Reynolds’dimensionless number. To obtain this value adimensionless parameter is used which is calledReynolds’ dimensionless number and flows more than2,300 Reynolds number were considered as turbulencestreams according to the environmental distribances.= . . = . = .. (1)

Where:V is the velocity of the fluid, D is the diameter of thepassage, is the viscosity, is the kinematic viscosity,

is the fluid density, Q is the volumetric flow rate andA is the area of the corssing surfa . In the computationalfluid dynamics (CFD), many attempts have been done

to define a comprehensive model for turbulence and thepossibility of this modelling has been provided for theusers in a variety of software environments.

A. Geometry productionFor production or draw geometries in Gambit the point,line, page and size commands are used. First thegeometry is drawn through pointing in the coordinatesystem and then by linking the lines the angles of thegeometry is drown. Finally a page is generated based onthe area created by the sides. In order to make thegeometry three dimensional the volume option is usedto link the pages. Thus the geometry of the componentswill be as follows. After completion of the geometry ofthe shape will be based on the following figure:

Fig. 2. The final shape of the geometry in the Gambit.

Hadad and Ghaderi 386

The boundary conditions include wall, the inflows, theoutflows, conditioning and etc.

DISCUSSION AND RESULTS

A. The results of the FLUENT programIn fact, this application is related to the mechanics usedin multi-phase flows. In this thesis the two phases of airand fluid are used. This software provides many resultsbased on various parameters but the results mentionedin the two-phased flow of air and wastewater are asfollows:(i) The convergence iteration curve in obtaining thesolution(ii) Fluid velocity and air mixing curves and the displayof the velocity vectors and speed changes on xy(iii) The display of the viscosity vectors, viscositymixture curves of the wastewater and air and showingtheir mixture on x-y.(iv) The display of the absolute pressure, static pressure

and dynamic pressure and pressure variations on x-y(v) The display of the mixture density and air density,velocity vector through the density and the densityprofile curve on x-y(vi) The curves of volume fraction of air with the fluidand the volume ratio of the air and fluis phases, thecurve of volume changes, and the profile curve ofvolume ratio on x-y.The curve of the convergence or divergence too obtainthe optimal answer in the iteration happens when all thelines of velocity in x, y and z direction reach 0.001which indicated that the iteration boundary isconverged and the answer is optimal. In the simulationof this paper due to the weakness of the PC and the lackof capacity to handle the high volume of the programthe iteration has been carried out up to 1000. Thefollowing diagrams indicate that in 1000 iteration theamount of changes was reduced which seems enoughhere (Fig. 3).

Fig. 3. Convergence iteration up to 1000.

The display of the velocity distribution in the aerationtank is presented in Figure (4). Since the tank is full andwall are introduced in the Fluent in the inlet and theoutlet, therefore the velocity of water or wastewater isstatic but the factor that cause the mixture is the air thatcan not be displayed in the 3D tank.Considering the lack of the observation of the velocityin the 3D version in the static flow Figure (4)shows theflat display of the middle of the tank in which thechanges happen.

In Figure (5) resulting from the flat view of the tank itis observed that the air injection rate in the inlet ismaximum which equals 2.63e+2m/s. Due to injectingair in the fluid the velocity is gradually reduced. If thereare numerous aeration pipes the time of presence of thewastewater in the tank becomes minimum. In thisthesis, the number of aeration pipes is more than 1 butunfortunately the computer was incapable of handlingmore than 1 pipeline and the research is limited to only1 pipe.

Hadad and Ghaderi 387

Fig. 4. The velocity distribution in the tank.

Fig. 5. The velocity distribution in the Tank.

Fig. 6. The static distribution of air into the tank.

Hadad and Ghaderi 388

Fig. 6 displays the static air velocity vector. Thevelocity vectors are extended from the exhaust pointthat is considered for a certain width range. In thisfigure the static pressure is the aeration hole is5.07e+4pa.Fig. 7 displays the velocity distribution vector. Themagnification of it is presented in the upper side of thefigure. At the beginning of the air outlet in the holes ofthe aeration pipe it is shown as the compact vector withthe maximum velocity.The profile of the velocity vector is shown in x-ydirection in Fig. 8 which consists of two diagrams. The

upper diagram is the flat diagram of the tank whichpresents the aeration velocity in each aeration hole thatpasses through the aeration pipe for the iteration 1000.The velocity in the inlet which is located at the centerof the coordinate system is higher and it is reducedgradually in the holes at the end of the pipe. The lowerdiagram presents the velocity in the whole tank. In thisdiagram it is observed that the velocity around the wallsand above the tank is zero which confirms the problem.The velocity is 2.5e+2m/s around the aeration holes.

Fig. 7. The velocity vector distribution.

Fig. 8. The velocity profile in the aeration and tank.

Hadad and Ghaderi 389

Fig. 9 shows the absolute pressure. According to thefigure the pressure is higher around the aeriation holes.The minimum pressure distribution in the aeration holesis 1.01 and the maximum pressure distribution inthe aeration holes is 2.32 . Wherever the locationand environment do not interfere the pressure, thepressure is the absolute pressure.

The static pressure in Fig. 10 indicates that where theaeration is higher the static pressure is lower and wherethe aeration becomes lower due to the waste the staticpressure will be higher. In this figure, it is observed thatthe maximum static pressure in the first aeration holesis 3.79 and 1.84 in the last whole whichindicates the maximum static pressure.

Fig. 9. The absolute pressure.

Fig. 10. The static pressure.

The dynamic pressure is the amount of drop within thepath presented in Figure (11). Wherever the movementof the air or the air velocity is high the dynamicpressure is high and wherever the movement of the airor the air velocity is low the dynamic pressure is low aswell. It has been observed that the dynamic pressure atthe beginning of the pipe is higher than its end andcovers a wider space in the tank. The highest dynamicpressure is at the beginning of the air injection which

equals 4.73 after mixing the two phases of airand water it is reduced around the aeration pipe andbecomes4.98 . Finally the dynamic pressurearound the walls and above the tankbecomes4.09 .The static pressure is obtained by the dynamic pressurebased on the following equation:Dynamic pressure= + (2)

Hadad and Ghaderi 390

Fig. 11. The dynamic pressure.

Dynamic pressure vector variations in the x-y directionare presented in Figure (12) which includes two shearsection pressure and the total tank pressure. It can beobserved that in the section of aeration pipe in the firstdiagram the dynamic pressure is higher at the beginningof the pipe compared to the end.The second diagram presents the dynamic pressure ofthe tank which indicates that the wall, inlet and theoutlet pressure is zero considering the wall in theboundary conditions but there is dynamic pressure inthe pipes and aeration holes the value of which is

presented. Obviously, the pressure in the aeration inletis higher than the outlet.Fig. 13 analyzes the aeration tank in the two phases ofwater and wastewater. Considering the results obtainedfrom this figure the density is lower than the tank at theair inlet and fluctuate between 6.06 / and1.20 / . It is concluded from this figure that thenumber of holes is less than the ideal condition becausethe air- wastewater mixing is not done optimally andthis mixture does not affect the tank and it has acted asa mass.

Fig. 12. Dynamic pressure at the aeration and the whole tank in x-y.

Hadad and Ghaderi 391

Fig. 13. The density of air and wastewater mixture.

Fig. 14 displays the density of air. It is observed thatthe mixture is not accomplished and due to the lack ofthe aeration pipes in various areas this result seemsrational. According to the figure the density of blue,green and yellow colors is equal everywhere the valueof which is 1.23 / and they vary in smalldecimal digits which depends on the accuracy of thecalculation and the lack of aeration pipes. The profile

density is presented in Fig. 15, the first diagram showsthe air density curve in the mixing mode in the B-Bsection and the second diagram shows the density curvein the whole tank.It can be seen that the density of the sides of the tankand the walls is 1.20 /m3 which is consistent withFig. 15.

Fig. 14. The density of the air in the tank.

SUMMARY OF THE RESULTS

(i) Fluent program has a specific ability of simulation inthe static flow in a closed structure in two different(biphasic) fluids.(ii) The changes of velocity in the two environments ofaeration and wastewater through the aeration pipes arepresented in the form of graphics within the tank.

(iii) Due to the lack of aeration pipes and the holesrelated to the aeration the system does not have thecapacity of optimal mixing with the air.(iv) At the level of air- wastewater mixture, generallydue to the lightness of the air in the whole system it canbe observed in the long term but a longer time shouldbe considered.

Hadad and Ghaderi 392

(v) The Fluent program is capable of presenting thechanges in velocity and pressure at different times andany section of the structure.(vi) Due to the lack of access to accurate information ofthe treatment plants and their lack of cooperation

generally simulation of the wastewater system is ademanding process but the Fluent program in case ofhaving access to computers with a powerful processorcan perform the simulation at different forms andvelocities.

Fig. 15. The air density profile curve in x-y.

REFERENCES

Nadafi. K, (2011). "Wastewater Treatment",construction and training organization, theMinistry of Energy, Iran.

Afuoni. M, Erfanmanesh. M, (2012). "EnvironmentalPollution" Arkaneh Danesh Publications, Isfahan,Iran.

Tovrovski, E.S. Maathai, P.K. , (2010). "Treatment ofSewage Sludge", Mousavi. GR (Translator),Hafeez Publications, Iran.

Bedelians Golikandi. G, (2012), "WastewaterTreatment Plant Designing", Aeezh Publication,Tehran, Iran.

Metkaf and Eddy, (2011), "Wastewater Engineering",Markazeh Nashre Daneshgahi Publishing, Design

and Research Consulting Engineers Company ofIsfahan Water and Wastewater, Iran.

Gourjar, B.R. (2012), "Sludge Treatment andDisposal", Mokhtarzadeh Azar.R (Translator),Hafeez Publishing, Rural Water and WastewaterCompany of West Azerbaijan Province, Iran.

Monzavi.MT, (2011), "Municipal Wastewater", TehranUniversity Publications, Iran.

Sarafraz, S., Khani M.M. and Yaghmaeian K. (2007).Quality and quantity survey of hospitalwastewaters in Hormozgan province. Iran. J.Environ. Health. Sci. Eng., 4(1): 43-50.

Youngs DL (1982). Time-dependent multi-materialflow with large fluid distortion. Num. Meth. forFluid Dynamics. Academic Press.

Hadad and Ghaderi 393

Wahl TL, Repogle JA, Wahlin BT and Higgs JA(2000). New developments in design andapplication of long-throated flumes. JointConference on Water Resources Engineeringand Water Resources Planning andManagement. Minneapolis.

Faure J.-B, Buil N and Gay B (2004). 3-D modellingof unsteady free-surface flow in open channel. J.Hydraulic Res. 42(3): 263–272.

Sarkar MA and Rhodes DG (2004). Calculation offree-surface profile over a rectangular broad-crested weir. Flow Measurement andInstrumentation, 15: 215–219.

Chen Q, Dai G and Liu H (2002). Volume of fluidmodel for turbulence numerical simulation ofstepped spillway overflow. J. Hydraul. Eng.128(7): 683–688.

Tokyay, N.D., “Effect of Channel Bed Corrugationson Hydraulic Jumps”, Impacts of Global ClimateChange Conference ,EWRI, 15-19 May,Anchorage, Alaska, USA, 2005.

Esmaili, K., Naghavi, B., KouroshVahid, F. andYazdi, J. (2010). “Experimental and numericalmodeling of flow pattern on circular weir”,Journal of Water and Soil, 24(1): 166-179.

Karim, M.F., Tanimoto, K., Hieu, P.D., (2009).Modelling and simulation of wavetransformation in porous structures using VOFbased two-phase flow model. Appl. Math.Model., 33: 343–360.

Taylor, E.H. (1944). "Flow characteristics atrectangular open-channel junctions",Transactions, ASCE, Vol. 109, pp. 893–902.

Weber, L. J., Schumate, E. D., and Mawer, N. (2001)."Experiments on flow at a 90◦ open-channeljunction.", Journal of Hydraulic Engineering,ASCE, Vol. 127, no. 5, pp. 340-350.

Huang, J., Weber, L.J., Lai, Y.G. (2002), "Threedimensional numerical study of flows in openchannel junctions.", Journal of HydraulicEngineering. ASCE, Vol. 128, no.3, pp.268-280.

Chong, N.B. (2006), "Numerical simulation ofsupercritical flow in open channel." M.S.Cthesis, University Technology Malaysia.

Pirzadeh, B. and Shamloo, H. (2007). "Numericalinvestigation of velocity field in dividing openchannel flow." Proceedings of the 12th WSEASInternational Conference on appliedmathematics, Cairo, Egypt, December 29-31,2007, pp.194-198.

Namaee MR, Rostami M, Jalaledini S, Habibi M. A 3-Dimensional numerical simulation of flow over abroad-crested side weir. Advances inHydroinformatics 2014; pp: 511-523.

Ramamurthy J., Tadayon A. S. R., Chen Z., 2009:Numerical simulation of sharp-crested weirflows. Canadian J. of Civil Engg., 36, 9, 1530–1534.

FLUENT user’s guide manual-version 6.3., FluentIncorporated, Lebanon, N.H., 2013.

Neary, V.S. and Odgaard, A.J. (1993), "Threedimensional flow structure at open-channeldiversions. ", Journal of Hydraulic Engineering.ASCE, Vol.119, no.11, pp.1223- 1230.

Recommended