Evaluation of CSO impact on receiving water quality in Tokyo coastal waters

Transcription

1 Japan-Korea Special Workshop on Impact assessment and ontrol of ombined Sewer Overflow Associated Event of the 4th IWA-ASPIRE onference & Exhibition, 2 October 2011 Evaluation of SO impact on receiving water quality in Tokyo coastal waters Hiroaki Furumai,, Keisuke Kojima, Akihiko Hata, Ikuro Kasuga, Futoshi Kurisu and Hiroyuki Katayama Research enter for Water Environment Technology, Department of Urban Engineering Graduate School of Engineering, The University of Tokyo 1

2 Safe and confortable water front for water- amenity in coastal area of Tokyo Bay A Odaiba seaside park A B Kasai Wakasu B Jyonan-jima seaside park Ohi-pier central seaside park Tokyo Bay 2

3 Odaiba Seaside park A Tokyo Bay Area: 50.7ha (Land 7.2ha Sea 43.5ha) Average Depth: 3m Volume: 1.40Mm 3 (at ebb tide) No. of Visitors: 1.67 Million in

4 Background (SO events in Tokyo: about 30 times per year) ombined Sewer Overflow (SO) In rainfall events, untreated sewage is often discharged into receiving water with dilution by rainwater ombined sewer system Separate sewer system BOD loading by SO accounts for about 40% of total loading on receiving water In receiving waters, there is concern about SO impact on health risk and hygiene issue at water front with water amenity activities SO outlets SO pump station Ref. Ministry of Land, Infrastructure and Transport In Tokyo, there are about 800 SO outlets and many pumping stations along urban rivers and in the coastal area. 4

5 Research Purposes To investigate the distribution of pathogenic indicator, Escherichia oliform(e coli.), at a coastal region of Tokyo Bay after a SO event To develop a numerical model describing the behavior of E coli. and to calibrate the model with the monitoring data Monitoring is necessary to understand the impact of the SOs on water pollution at a seaside park Numerical model for the behavior of E coli. would be helpful to evaluate pathogenic pollution level quantitatively. alibrated model can be used to predict pollution level and to identify the most important pollutant loads. It is also applicable to explore countermeasures for water pollution control and to evaluate their effectiveness. 5

6 Extensive Monitoring 6

7 Extensive Monitoring Long term monitoring after SO event To observe the duration of pathogenic pollution at the coastal area caused by SO Sea water level (m) Sampling Total precipitation: 27 mm Maximum hourly precipitation: 3.5 mm 7 sampling times Nov. (date) Precipitation [mm/hour] 7

8 Monitoring points and analysis of coliforms To investigate distribution of pathogenic indicators at 5 points (river inflow point and St.1-St.4 in coastal area) Sumida river inflow point St. 1 Sewage pumping station W1 St. 2 St. 4 St. 3 W2 WWTP (Waste Water Treatment Plant) Monitoring point Odaiba seaside park W3 1 km E. coli and Total coliforms were simultaneously determined by a single agar layer method using hromocult coliform agar (Merck). 8

9 Distribution of E. coli concentration (Surface water) 11 November (Ebb (Flood 12 November tide) (Ebb (Flood tide) tide) 14 November (Ebb tide) St. 1 St. 2 W2 Sumida river inflow point Sewage pumping station W1 St. 4 WWTP (Waste Water Treatment Plant) 21 November (Flood tide) St. 3 Monitoring station 28 November (Ebb (Flood tide) tide) W3 1 km [FU 100mL -1 ] 9

10 Temporal change of E. coli oncentration Variation between ebb and flood tides Surface water Bottom water Detection Limit E. oli conentration [FU 100mL -1 ] Detection Limit Surface water Bottom water St. 1 St St Surface water Bottom water Detection Limit November W1 St. 2 W3 St. 3 St. 1 St W2 1 km

11 Summary (Extensive Monitoring) The extensive monitoring showed clear evidence of fecal contamination caused by SO at a seaside park as well as the river inflow streamline. Two types of temporal variations in the E. coli concentration were observed. One is variation between ebb and flood tides and another is downward movement of polluted water mass by river inflow. Regarding to the downward movement of highly polluted water mass, there was a time lag in increase of E. coli concentration found at downstream station (St. 2, St. 3). The monitoring data showed that it took four days to decrease the contamination level to the bathing water quality criteria at St. 1, St. 2 and St. 4, while St 3 had higher contamination than other points 4 days after the event. 11

12 3-Dimensional simulation 12

13 3-Dimensional Flow and WQ Model Input data Bathymetric feature Meteorological Data air pressure, vapor pressure, temperature, relative humidity, wind speed, wind direction, cloud amount, solar radiation, precipitation River discharge (inflow) Open boundary data such as sea level, water temperature, salinity Sewage (SO) discharge at the pumping stations and Effluent discharge at the wastewater treatment plant Model structure 3Dimensional model (σ-coordinates) Navie-Stokes equation Mass balance equation with advective diffusion, settling and decay of E. coli Output data Water velocity, sea water level Water temperature, salinity E. coli conc. 13

14 Bathymetric feature Tokyo Bay scale oastal scale 0 100m 0 20m 14

15 3-Dimensional Simulation Area 100m-grid St. 1 Sumida river inflow point Odaiba seaside park 2000m-grid W1 St. 2 St. 3 St. 4 W2 Sewage pump station WWTP (Waste Water Treatment Plant) Monitoring station Nesting at the open boundary 10 km W3 1 km Sea water level Water temperature Tokyo Bay Scale oastal Scale Salinity 15

16 Advection Advection-diffusion equation diffusion equation (E. coli E. coli) Settling velocity ( E. coli inflow load to the upper layers ) V outflow outflow flow F K R F R z K z y K y x K x z W w y v x u t W W P P R R ex D D ex D Z Y X ) / ( ) ( ) ( ) ( ) ( + + = = = Decay Upper layers River discharge and SO discharge 16 16

17 onditions of river flow and runoff discharge River flow rate was summation of base river flow and runoff discharge. The runoff discharge was calculated by unit hydrograph method with rational formula depending catchment area and travelling time. Hydrograph method with rational formula 1 Q = I A 360 : runoff coefficient I : rainfall intensity A : catchment area Discharge (m 3 /s) a b b a c 足し合わせ Sum c Intensity (mm/hr) hour 17

18 ondition of E. coli concentration from river E. coli concentration change in river inflow was estimated based on the monitoring data at 1km upstream from the river boundary grid. The background E. coli concentration was given as 100 FU /100mL in river flow and coastal water before the rainfall Backgroun d E. coli level E. coli [FU/100mL] E. coli concentration in river Precipitation Measured value (1 km upstream from river flow input point) 1.0E E E E E E E E E R = t 3.56 riteria of E. coli for bathing t (day) November Precipitation [mm/hour] 18

19 ondition of E. coli concentration from Pumping Station (p) and WWTP (w) P W = = S sewage flow rate sewage and rainwater flow rate W secndary treatment flow rate + WP primary treatment secndary and primary treatment flow rate flow rate S, W, WP are E. coli concentrations in dry weather sewage, secondary effluent and primary treatment discharge. Sewage flow rate = dry weather flow rate (Q) changing with time using variation coefficient. Runoff (rainwater) flow is estimated by unit hydrograph and rational formula depending on drainage area and travelling time. Primary treatment discharge: intercepting flow rate (2-3Q) secondary treatment flow rate (dry weather flow=q) 19

20 Simulation results of E. coli concentration (Surface water) Sea water level (m) Precipitation [mm/hour] Nov. (day) 0 8 FU / 100mL 20

21 Simulation results at St. 1, Measured value (Surface layer), Measured value (Bottom layer) alculated value (Surface layer) alculated value (Bottom layer) E. coli [FU/100mL] 1.0E E E E E E E+02 Precipitation Surface water Bottom water St. 1 (Depth=7.8 m) riteria of E. coli for bathing Precipitation [mm/hour] 1.0E E November 21

22 Simulation results at St. 2, Measured value (Surface layer), Measured value (Bottom layer) alculated value (Surface layer) alculated value (Bottom layer) E. coli [FU/100mL] 1.0E E E E E E E+02 St. 2 (Depth=10.6 m) riteria of E. coli for bathing Precipitation [mm/hour] 1.0E E November 22

23 Simulation results at St. 3, Measured value (Surface layer), Measured value (Bottom layer) alculated value (Surface layer) alculated value (Bottom layer) E. coli [FU/100mL] 1.0E E E E E E E+02 St. 3 (Depth=10.4 m) riteria of E. coli for bathing Precipitation [mm/hour] 1.0E E November 23

24 Simulation results at St. 4 (Odaiba seaside park), Measured value (Surface layer) alculated value (Surface layer) alculated value (Bottom layer) E. coli [FU/100mL] 1.0E E E E E E E+02 Precipitation Surface water St. 4 (Depth=3.0 m) riteria of E. coli for bathing Precipitation [mm/hour] 1.0E+01 Bottom water 1.0E November 24

25 Summary (3-dimensional simulation model) Spatial distribution and temporal changes of E. coli concentration were simulated using 3-dimensional 3 water quality model. This simulation suggested that E. coli loading from the river inflow can have a greater impact on increasing trend in E. coli concentration in coastal area including the Odaiba seaside park. However, the simulation results overestimated the E. coli concentration in surface water during several days, while the concentration was relatively well-expressed expressed in the bottom water. It is necessary to elaborate the input condition of pollutant load from river inflow so that model calibration could be sufficiently achieved for E. coli behavior prediction in the coastal area. 25

26 Points to be improved Grid resolution? River inflow load condition? Background E. coli level E. coli [FU/100mL] 1.0E E E E E E E E E R = t 3.56 riteria of E. coli for bathing t (day) Precipitation [mm/hour] November 26

27 Thank you for your kind attention 27