DEVELOPMENT OF TWO-DIMENSIONAL HYDRODYNAMIC AND WATER QUALITY MODEL FOR HUANGPU RIVER DEVELOPMENT OF TWO-DIMENSIONAL HYDRODYNAMIC AND WATER QUALITY MODEL FOR HUANGPU RIVER

DEVELOPMENT OF TWO-DIMENSIONAL HYDRODYNAMIC AND WATER QUALITY MODEL FOR HUANGPU RIVER

  • 期刊名字:水动力学研究与进展B辑
  • 文件大小:865kb
  • 论文作者:XU Zu-xin,Yin Hai-long
  • 作者单位:State Key Laboratory of pollution Control and Resource Reuse
  • 更新时间:2020-07-08
  • 下载次数:
论文简介

Journal of Hydrodynamics, Ser. B.2 (2003),1- 11China Ocean Press,Beijing 一Printed in ChinaDEVELOPMENT OF TWO-DIMENSIONAL HYDRODYNAMIC AND WATERQUALITY MODEL FOR HUANGPU RIVERXu Zu-xin, Yin Hai-longState Key Laboratory of pollution Control and Resource Reuse, Tongji University, Shanghai 200092,China( Received Nov. 27, 2002 )ABSTRACT: Based on numerical computation model RMA2 serious pollution to Huangpu River, so the waterand RMA4 with open source code, finite element meshes re-quallity issues are primarily concerned.presenting the study domain are created, then the finite ele-In the year of 2000-2001, a two-dimensional fi-ment hydrodynamic and water quality model for Huangpunite element hydrodynamic and water quality modelRiver is developed and calibrated, and the simulation resultsfor Huangpu River was developed so as to lay theare analyzed. This developed hydrodynamic and water qualitymodel is used to analyze the influence of dischargedscientific foundation for making “the tenth -fivewastewater from paning Wastwater Treatment Plant year (2001-2005) plan on Shanghai's water envi-( WWTP) on Huangpu River's water quality.ronment protection and rehabilitation. ”Under thisplan, a wastewater treatment plant will be built a-KEY WORDS:Huangpu River water quality. finite element long the lower reaches of Huangpu River. Empha-method, Wastwater Treatment Plantsis would be laid on the impact of planning WWTPupon the water quality of Huangpu River, because :the comprehensive development project in the place1. INTRODUCTIONHuangpu River,originating from the Taihu,near the North Bund along two sides of the Huang-pu River has been launched andthe planningruns across Shanghai and feeds into Y angtze RiverWWTP is not far away from this place. The devel-at W usongkou. It is a tidal influencing river withoped hydrodynamic and water quality model hasmore than 50 tributaries among which some majorbeen used to carry out the evaluation of impacts oftributaries are shown in Fig. l.planning WWTP upon the Huangpu River.2.SETUP AND CALIBRATION OF TWO-DI-YunzaobartgMENSIONAL HYDRODYNAMIC MODEL Scatter points in Map bour.MYHC N M H Gesn't vary obviouslyModule of SMS) is used to form XYZ file of SMS,over space scale, so quadrilateral meshes are quali-from which data sets representingx, y and z coordi- fied for representing the study domain, triangular .nates are obtained. As a XYZ file shown in SMS meshes are not adopted here. For the whole study(Fig. 3),for a selected survey point, x coordinate area, each mesh is set as about 200m length in lon-is 25138.们存数握oordinate is 32272. 23m, z coordi- gitudinal dreection and 50m width in lateral directon. The study domain is divided into 7 meshes inlateral direction.Automatic and hand mesh gererating schemesare combined to form the meshes. First, the wholestudy area is cut into a series of sub-area; for eachsub area, the four boundaries of the area is speci-fied by hand. Second, the meshes are generated au-tomatically in each sub-area. Some created meshesare shown in Fig. 4. it can be seen from Fig. 4 themeshes are aligned with the flow direction and theyare all quadrilaterals. For the whole study domain,the number of elements is 2969,the number ofnodes is 9768.Fig.5 Values of bathymetric isolines of study domain inmeters2.6Initial and Boundary ConditionsInitial Conditions: the influence of initial con-ditions will disappear after several tidal cycle hoursof computation, so the initial conditions are thesame elevation of water surface and zero velocity ilall nodes in the study.Fig.4 Boundary fitted meshes for Huangpu River ( nearBoundary Conditions: Huangpu River is ariver moush of Suzhou Creek)strong tidal influencing river, the flow and watersurface elevation are mainly influenced by tidal wa-2.5 Survey data interpolation to mesh nodester from mouth of Yangtze River, in contrast theIn the process of mesh generation, the location flow from tributaries is influenced much less, soof each node can be assigned automatically, howev- the boundary conditions about tributaries can beer,the bathymetry of each node can not be as- excluded from consideration. in this model, twosigned, so scattered survey data sets (as shown in boundary conditions are speified, that is, the up-Fig.2 and 3) are used for interpolation to nodes ofstream discharge or water level boundary conditionfinite element to help construct the bathymetry. A- in Mishidu and downstream water level boundarymong several interpolation techniques, inverse dis-condition in Wusongkou. Because tidal water cantance weighted interpolation technique is used; it isgo upstream to Mishidu, both the upstream andbased on the assumption that the interpolation sur-downstream boundary condition aredynamicface should be infuenced most by the nearby pointsboundary conditions.and less by the more distant points. The interpola-2.7 Calibration of hydrod ynamic modelting surface is a weighted average of the scatteredCalibration of the model requires considerationpoints and the weight assigned to each scatter pointf both the discharge and water level to adjustdiminishes as the distance from the interpolationManning's coefficient. In our study, the synchro-point to the scatter point increases.According to this interpolation technique, thenous hydraulic monitoring data from May 26th tobathymetry of each node were assigned. In theAugust 29th of 1999 are used to calibrate the mod-calibrated Manning' swhole study area,the minimum bathymetry is中国煤化工coeffor tital flow, the cali-29.69m,the maximum bathymetry is - - 0. 69m.brate.THCNMHGis0.020.Anotherpa-F urthermore the bathymetric isolines can be made.rameter is Peclet number which represents the rela-Fig.5 shows the bathymetric isolines of study do-tionship between velocity and eddy coefficient, andmain near river mouth of Suzhou Creek.is defined as:5Table 1 Comparison of simulated and measured results (time series: 1999. 6.3-1999. 6.17)hydrologicrelative error percent covering total data sets (total data sets are 2160) standardconstituentslocation10%15%20 %30 %40%<5%> 50% error-15%-20%-30 %-40%-50%Songpu Bridge99.58 0. 420o. 037sectiontidal waterDianpu River84.44 15. 560.082levelmouth sectionW usongkou1000. 00151.0615. 975.23 4. 91 .6.253. 983.159.45 162. 21tidal flow31.29 10. 1917.13 10.37 3.615.276.85 15. 29498. 4525.30 6. 396. 656. 3922.6912. 57.3618674. 78Songpu Bridge section3.5P=pud.x(4)器2.5|Esimulatedmeasuredwhere P -Peclet number, ρ一fluid density;0 4896144 192240288336384u一- average elemental velocity, dx-length of6000mouth of Dianpu River sectionelement in stream-wise derection; E -:ddy vis-cosity. Peclet number is set as 10-20 during compu-tation. Results of callibration can be seen in Fig. 6.2 -000For model calibration, comparison of measured-6000 ;04896144192402888336384and simulated results is shown in Table 1. The se-- - - simulated“n... measuredlected locations are Songpu Bridge section, DianpuRiver Mouth section and W usongkou section inHuangpu River where tidal level and flow data aremonitored. Two indexes are used to examine theFig.6 Simulated and measured results of two selectedcalibration results, that is, relative error andsections in Huangpu Riverstandard error.mearThe fllowing conclusions can be drawn from O. 1中国煤化工' thanmodel is qualified toTable 1:pred:MYHCNMHGof Huangpu River at(1) The hydrodynamic model can simulate pre-any time.cisely the tidal water level at any location in Huan-(2) The hydrodynamic model can be put intogpu River. In the selected 3 locations, the numberpractice in simulating tidal flow of any locaton inof data:sH数hich the relative error is less thanHuangpu River. In the selected 3 locations, the10% covers 90%0 or more of the total date sets, thenumber of data sets in which the relative error is erning equation: .less than 30% covers 70% or more of the total datasets,and the nean- square error is not large com- (HC)」 (uHC) 」A(vHC)_ a( HD,aCpared with the flow. The simulated tidal flow from)tayJxdx'this model is satisfactory, because the measureddata also differs from actual flow due to some fac-(HD, )+S;+S。(5)dytors such as the error of lateral cross section infor-mation and measured data.where Cconcentration of state variable of in-。2Wusongkou sectionterest; u, V-depth-averaged velocity in the Car-tesian direction; Hwater depth; D,,D,disperision coefficient in the Cartesian direction; S-source/sink item; S;- source/ink term fromkinetic biological and chemical processes. u, v,H0 48 96 144 192240288336384come from hydrodynamic simulation results./hFor each state variable, only term from kineticprocess S; differs. The kinetic equation describingeach state variable is:Fig.7Simulated tidal velocity of selected location inHuangpu River(1) C= COD,dCOD -=- k.CoD;dt(2) C= BOD。,dBOD -- k,BOD;Ht(3) C= NH3- N,dNH3- N=- k。NH3-N;d(4)dDO= k2(O,-O)-km BOD3 -4.57k。NH3- N- SOD.where kCOD decaying coefficient; kBOD: decaying coefficient; k, - NH3-N decayingcoefficient; k;- reaeration coefficient; SOD-Fig. 8 Simulated tidal currents during low and high tidesediment oxygen demand. O, is saturated concentra-(left: low tide, right: high tide)tion of dissolved oxygen. Regardless of salinity in-(3) The tidal velocity can also be given in thefluence onO,,O, only has relationship with temper-process of simulation. Fig. 7 shows the time seriesature T. The following equation is adopted for es-of tidal velocity in W usongkou section during calli-timatingO, :bration. Fig. 8 shows the tidal currents during lowO, = 14.55-0.3822* T +0.005426* T° (6)and high tide of calibration. Fig. 7 shows the maxi-mum high tide veloeity can rech. 1. 8m/s. this re- 3.2 Imitial and boumndary cnditionssult matches well with the recorded velocity data,Initial Conditions: The initial conditions areso it is believable and the simulated velocity ( veloc-formed after computation of several tidal cycles,ity component u and U ) can be accepted as an inputthat is, a hot-start input file is made.to water quality model for Huangpu River.Boundary Conditions: The upstream concen-tration boundarv. condition in Mishidu and down-3.SETUP AND CALIBRATION OF TWO-DI-strea中国煤化工ry condition in Wuso-MENSIONAL WATER QUALITY MODELngkdHCNMHGtothemonitoringre-3.1 Governing equationssults from 1998 to 2001.In this study, the two-dimensional water quali-3.3 Estimate of pollutant loadsty model for Huangpu River that is capable of mod-In the old water quality model for Huangpueling the_ state. variables COD, BODs, NHs-N andDO is de厄存貉据It is based on the following gov-7Table 2 Comparison of simulated and measured results (time series: 1999. 8.19-1999. 8.24)relative error percent total data setsstandard .variablelocation0-10%10-20% 20%-30% 30%-40% 40%-50% > 50%errorSongpu Bridge section4.314.351.415.703.28CODLinjiang section10.7249.317.310. 7Yangpu Bridge section50. 631. 2133.91.3C3.1751.237.29.00. 40BOD;11.514. 138.523. 17.75.11.0228.628. 620. 69.57.94.81. 3320.524.416.714.112. 80.41NH3-N .15.620.832.519.59.12.60.4528.2.39. 76.466.728. 25. 10.3832.110.20.63.Y angpu Bridge section25.63.821.80. 47River,the pollutant loads to the model are esti- 1999. The two-dimensional hydrodynamic model ismated by the following way: the Huangpu River is used to simulate u,U, H as input to water quality .divided into a serious of segments in the longitudi- model. During this period, the average temperaturenal direction (for example, one segment per 2km),is 29-30°C,and the decaying rate of the state varia-a border line (for example, 2km from the geomet- bles varies with temperature. The equation used torical boundary of Huangpu River) is established as describe decaying rate of sate variables under differ-the boundary of area for estimating pollutant loads, ent temperatures in RMA4 is .the pollutant load in each sub-area (2 X 2km) issummed as the total pollutant loads to the model ofk(T) = k(20)●(1.047)(T-20)(7)each segment. This method can be applied in rivernetwork,however, it is not suitable for Huangpuwhere k(20) is decaying rate in 20°C, k(T) is deca-River model, for the pollutant loads generally gying rate in temperature Tinto Huangpu River's large tributaries first, thenThe calibrated coefficients is listed as follows;reach Huangpu River with flow of tributaries. So aE= 50m2/s, k。=0. 10-0.12d-', kr = 0.16-practical method is to estimate the pollutant loadsinto its major tributaries, and treat the pollutant0.20d-',k,=0.07-0.09d-1,k2=0.20d-', SODloads form the large tributaries as point sources fo中国煤化工In addition, in theHuangpu River.proc:JYHCN MHG'within the range of3.4 Calibrationof the modelcalibllutant loads can alsoThis two dimensional water quality model was be adjusted, thus the pollutant loads into Huangpucalibrated on the basis of water quality monitoring River during various seasons of each year can be es-data from Shanghai Environmental Monitoring Cen- timated. Simulated and measured COD and BOD; atter in the p互執据f August 19th to August 24th of location of Songpu Bridge section can be seen inFig. 9, which shows the simulated results are rea- techniques; (3) The velocity computed from hydro-sonable. COD concentration distribution over the dynamic model is accepted as input to water qualityspace of study domain is shown in Fig. 10.model, the error of simulated velocity also exists.Even the precision of simulated results is limit-30pYangpu Bridge sectioned by the above facts, however, this water qualitymodel can analyze the trend of changing state varia-g10bles with tide successfully, so it can be put intosimulated● measuredpractice in water quality prediction.24487296120144Yangpu Bidge secion4APPLICATION OF THE HYDRODYNAMICAND WATER QUALITY MODEL4.1 Backgroundg 4tAccording to“the tenth-five year (2001 -2005)24 4{72元96 120 144plan on Shanghai' s water environment protectionand rehabilitation”, a waste water treatment plantwill be built in north of Fuxing Island along lowerFig.9 Simulated and measured results of selected sec-reaches of Huangpu River, just as Fig. 11 shows.The designed treatment capacity is 500000m2 /d.tions in Huangpu RiverThe peak flow in dry seasons is 7. 37m2/s,the av-Wusongkouerage flow in dry seasons is 5. 79m3 /s,the overflow2sK30H3discharge in rainy days is 21. 0m3 /s. The specifiedwater quality standards of tail water from WWTPmouth of Suzhou CreckYangpu Bridgeare: COD,120mg/L; BODs, 30mg/L; NH-N,10mg/L. This is a very important wastewaterLinjiang\(20treatment project during the tenth-five year whichwill exext great influence on Shanghai's water enSongpu Bridgevironment rehabilitation.|. wy 15MishiduHowever, because this WWTP is not far awayFig.10 COD concentration distribution over the space offrom the planning EXPO 2010 exhibition place andstudy domain during maximum high tide calibra-location for comprehensive launching project alongthe two sides of Huangpu River's lower reaches thetion time series .WWTP's impact on Huangpu River's water quali-ty must be evaluated carefully, the two-dimensional3.5 Analysis of simulation resultsComparison of simulated and measured results .hydrodynamic and water quality model is herebyis shown in Table 2. It can be seen from Table 2applied.4.2 Pollutant loadssimulated COD is more satisfactory than other sateAfter the WWTP is completed, the pollutantvariables, this is because COD is not coupled withloads which flow into Hongkougang and Yangshu-DO and can be simulated independently, so its cali-pugang previously will be sent into WWTP (shownbration is not difficult. In contrast, BOD and NH3-in Fig. 1l),thus the pollutant loads into Hongk-N are coupled with DO,so the difficulty of theirougang and Y angshupugang will decrease. Mean-calibration is large. Among 4 state variablesm, cal-while,because this collected wastewater will beibration of DO is the most difficult.treated in W WTP. the pollutnat loads into Huang-Compared with hydrodynamic model, calibra-pu中国煤化工general. By using GIStion of water quality model is affected by severalsoftvYHC N M H Ge pollutant loads intouncertain facts: (1) The pollutant loads vary at va-WWTP by wastewater collection system is CODrious time of each day in theory, but in practiceon- 155.5 tons/d; BOD 60. 5 tons/d; NH-N 10. 4ly daily or monthly pollutant loads can be offered;tons/d.(2) The error of field water quality data can't be a-voided du2方数梔limitation of available measuring9Table 3 Scenarios for evaluating influence of discharged wastewater on Huangpu RiveNoscenario(1) hydrologic condition of normal year (2) without WWTP(1) hydrologic condition of dry year (2) without WWTP(1) hydrologic condition of normal year(2) with WWTP, secondary treatment , side discharge, discharge flow 7. 52m* /s(2) with WWTP, secondary treatment, off -river discharge, discharge flow 7. 52m3 /s(1) bydrologic condition of dry year(2) with WWTP, secondary treatment , side discharge, discharge flow 7. 52m3 /s(2) with WWTP, secondary tretment, off river discharge,discharge flow 7. 52m2 /s(1) bydrologic condition of normal year(2) with WWTP, secondary treatment, side discharge, discharge flow 21. 68m* /s(2) with WWTP, secondary treatment, off-river discharge, discharge flow 21. 68m2 /s. (1) low upstream inflow, spring tide, typhoon(2) with WWTP, secondary treatment, side discharge, discharge flow 21. 68m3 /s(1) low upstream inflow, spring tide, typhoon(2) with WWTP, secondary treatment, off- river discharge, discharge flow 21. 68m2 /s .Table 4 Concentration field of COD due to wastewater dischargeincrement scenarioscenario scenario45781(> 10mg/L ./102. 6length of> 8mg/L188.1143.1pollution789.8310.2255. 5150.2zone> 4mg/L 43. 9195. 837.66576. 55781. 2989. 1793.6二> 2mg/L984.5380. 8107891069117564165292378221449.4> 10mg/L33. 726. 2width of67.453. 7pollution > 6mg/L /388. 3383. 2112.4115.6> 4mg/L17.3195.8.27.4529.4504.8526.2350.6(m)> 2mg/L 529. 1522. 9.529. 4524. 6532. 4534.6area of1. 82中国煤化工pollution> 8mg/LMHCNMHG5.49> 6mg/L .215.9.95. 4521. 2113. 56(X0. 629.490.753477. 12840. 8428296.4103 m2 )> 2mg/L381. 7172, 45762. 45626. 886648487. 211665. 310403. 810Table 5 Influence of discharged wastewater on water quality of typical location( scenario 9, worst design hydrologic condition)scenariolocation .state variableconcentration increment(mg/L)maximumminimumaverageCOD3. 711. 502. 54YangpuBOD;0.820. 270.53Bridge0.330. 110. 222. 660.501. 18Nanshi WaterBODs0.570. 080.23scenario 9PlantNHx-N0. 240, 030.101.500.200. 56LinjiangBOD,0.300.030. 10Water PlantNHs-N .0.140.010.050. 000.00SongpuNH3-Nwherew= 2π/T,T= 24.83h, Q。= 150m3 /s.As for the typhoon occurence in scenario 7 andMmtingwwte8, it is assumed tidal water level increases 0. 8m onthe basis of spring tide water level.4.4 Simulation of concentration field due to .wastewater dischargeThe intention of simulating concentration fieldis to analyze the influence of discharged wastewateron Huangpu River. In this study, the pollutionFig.11 Depiction of planning Minxing WWTPzone where concentration is above one specified val-ue is used to depict the concentration field. Table 4shows the concentration field of COD due to4.3 ScenariosScenarios for evaluating influence of dischargedwastewater discharge at the time of maximum highwastewater on Huangpu River are listed in Tabletide.According to Table 4,the following conclu-In the scenarios, off-side discharge is specified sions can be drawn:as center discharge. The low upstream inflow ir(1) Compared with normal year hydrologiocscenario 7 and 8 is calculated by the following equa- condition, the influence of discharged wastewatertion:中国煤化工uality is obvious in dryQ(t) = Q一281. 9sinct一1234. 5sin2ut-year:YHCNMHGtc)uivent saile wasiewater discharge and de191sin3wt - 221. 7cosut十signed hydrologic condition, compared with sidedischarge,area of pollution zone and maximum1442. 3cos2at十45. 3cos3wt(8)concentration increment in the near field decrease ifcenter discharge is adopted,the reason is listed as:11the flow field of downstream Huangpu River ap- wastewater from planning WWTP will not influ-pears in longitudinal direction, the discharged ence the water quality of drinking water sourcewastewater will be limited in narrow zone near west place under scenario of storm, typhoon, spring tideHuangpu River side if side discharge is adopted; in and low upstream inflow. .contrast,the wastewater will be mixed with watermore easily if center discharge is adopted. Compari- 5CONCLUSION REMARKSson of two discharge schemes is shown in Figs. 12,The developed two- dimensional finite element1, 2,3. Fig. 12 represent contour of 6mg/L, 4mg/ hydrodynamic and water quality model has been ap-L, 8mg/L respectively.plied to provide scientific foundations for the mak-ing“the tenth-five year (2001-2005) plan on Shang-hai's water environment protection and rehabilita-tion”. This model is also a great improvement topreviously developed Huangpu River model, it can .be used to analyze the tidal water level, flow,locity and concentration of state variables in Huan-gpu River. Later study will focus on the simulationof more water quality constituents such as eutroph-ication process.Fig. 12Concentration distribution of two differentwastewater discharge schemes (left is side dis-charge, right is center discharge)REFERENCESCompared with scenario 7 and 8,the length [1] xU Zuxin. Theory and practie of river polution reha-bilitation[ M]. Beiing: China Envoronmental Scienceand area of pollution zone increase obviously underPress, 2003. (in Chinese)worst designed conditions ( scenario 9 and 10),so[2] ZHANG Chao, CHEN Bingxian,WU Lunhe. Geome-scenario 9 and 10 are the most typical schemes to e-try information system [ M]. Beiing: Higher Educationvaluate the influence of discharged wastewater onPress,1995. (in Chinese) .drinking water source place. Table 5 lists simula-tion results of scenario 9, it shows discharged中国煤化工MHCNMHG

论文截图
版权:如无特殊注明,文章转载自网络,侵权请联系cnmhg168#163.com删除!文件均为网友上传,仅供研究和学习使用,务必24小时内删除。