5.3 Description of the Environment

5.3.1.1 Routine water quality monitoring is carried out by EPD at marine water sampling stations in Deep Bay and North Western WCZs. The Deep Bay WCZ includes three monitoring stations in the inner sub-zone (DM1, DM2 and DM3) and two monitoring stations in the outer sub-zone (DM4 and DM5). Based on EPD's monitoring results for 2000, water quality in the outer sub-zone was better than that in the inner sub-zone. The BOD5, SS and inorganic nutrient levels were comparatively higher in the Inner Deep Bay. There was an increase in DO (0.4 mg/L) in the bay. At DM1 and DM2, the recorded DO levels were the lowest (3.6 mg/L at DM1 and 3.9 mg/L at DM2). There were remarkable increases in E. coli levels at all marine monitoring stations in Deep Bay. The increases ranged from 40% to 400%. There was an overall 17% decrease in BOD5 level based on the data recorded at the 5 monitoring stations. The nitrogen and phosphorus levels did not have significant variations from the data recorded in 1999.

5.3.1.2 EPD also carries out routine monitoring of river quality in Hong Kong. A Water Quality Index (WQI) with grading from excellent to very bad is used to indicate the extent of organic pollution in a river. In the Deep Bay WCZ, there was substantial improvement in water quality of Tin Shui Wai Nullah in 2000. The compliance rate improved from 49% in 1999 to 88% in 2000. The WQIs of the upstream and downstream stations in Tin Shui Wai Nullah were improved to "good" and "fair".

5.3.1.3 There are six water quality monitoring stations (NM1, NM2, NM3, NM5, NM6 and NM8) in the North Western WCZ. Outfall discharges from the Pillar Point, San Wai and Siu Ho Wan Sewage Treatment Works affected the water quality in terms of E. coli and BOD5 in the regions near the discharge locations. Based on EPD's marine water monitoring results for 2000, the dissolved oxygen (DO) and unionised ammonia (UIA) levels complied with the WQOs at all the stations. Except for station NM5, all other stations met the WQO for total inorganic nitrogen (TIN). The increases in sewage discharge from the NWNT outfall into Urmston Road might cause the rise in E. coli levels. This was reflected in the data measured between 1998 and 2000. Under the influence from the Pearl River flow, there were increasing trends in TIN and orthophosphate phosphorus at NM5 and NM3.

5.4 Water Sensitive Receivers

5.4.1.1 The outfall location at Urmston Road is within the North Western WCZ. In case of emergency, the discharge may be released into the Deep Bay WCZ or North Western WCZ. The indicator points including water quality sensitive receivers that are potentially affected by the proposed project and EPD's marine water sampling stations are listed below:

Note: * not included in water quality assessment

5.4.1.2 Figure 5.1 shows the locations of the indicator points located in the Deep Bay and North Western WCZs. Due to the proximity of some sensitive receivers and the size of the adopted model grid, not all of the sensitive receivers listed in Table 5.1 are included in the assessment. Sensitive receivers that are excluded from the assessment are marked with '*'.

5.4.1.3 In the event of emergency discharge to Deep Bay, treated effluent or raw sewage would be discharged via Tin Shui Wai Drainage Channel. The water quality in this drainage channel would be temporarily affected. However, since the channel is not a natural channel and does not have high ecological values, it is therefore not included as one of the indicator points in the water quality modelling.

5.5 Assessment Methodology

5.5.1 Construction Phase Impact Assessment

5.5.1.1 Potential sources of water quality impacts are likely to be the construction site runoff, sewage from workers and canteens on the site and accidental spillage of chemicals. Water quality impact assessment in this section covers all the identified potential sources. Mitigation measures are proposed based on the ProPECC Note on Construction Site Drainage.

5.6 Operational Phase Impact Assessment

5.6.1 General

5.6.1.1 Upon completion of the proposed project, the treated effluent from the San Wai STW will be discharged to the North Western waters via the outfall at Urmston Road. In an emergency event, the treated/untreated effluent will be discharged to the Deep Bay waters or North Western waters (see Section 1.2 of this report). The prediction of water quality changes as a result of effluent discharge was made using mathematical modelling tools. The proposed water quality modelling methodology is presented in the following sections.

5.6.2 Far-field Model

5.6.2.1 The Delft3D suite of models was used to provide a platform for hydrodynamic and water quality modelling. A detailed model, which was developed and calibrated under the ongoing project "Review of North District and Tolo Harbour Sewerage Master Plans (Agreement No. CE 28/99)", was used for simulation of the hydrodynamic and water quality changes due to the upgrading and expansion of San Wai STW. This detailed model covers the North Western, Deep Bay, Victoria Harbour, Western Buffer and Eastern Buffer Water Control Zones (WCZs) and is direct applicable to this Study. The grid size of the detailed model was designed to range from 400m in open waters and less than 75m in the regions near Urmston Road, Brothers Point and Victoria Harbour Channel.

5.6.2.2 The detailed model was linked to the Update Model, which was calibrated and verified by Delft Hydraulics under the Agreement No. CE42/97, to provide boundary conditions. As the Update Model covered the whole of Pearl River Estuary, Macau, Lama Channel, Deep Bay and Mirs Bay, influences on hydrodynamic and water quality from these outer areas was fully incorporated into the detailed model. The Update Model also provided information on the initial conditions to the detailed model.

5.6.3 Modelling Scenarios

5.6.3.1 The time horizon for modelling is the 2016 scenario. Factors incorporated into the model include:

5.6.3.2 The flow and load data outside of the Study Area were based on the Pollution Load Inventory compiled under the study "Review of North District and Tolo Harbour Sewerage Master Plans (Agreement No. CE 28/99)". The pollution sources within the Study Area were updated to take into account the changes due to the expansion and upgrading of the San Wai STW. The final flow and load data were fed into the detailed model for prediction of the potential water quality impacts. Both dry and wet seasons were simulated in the modelling work for a 15-day spring-neap cycles. The model spin up covers 45 days prior to the actual model simulation period. The modelling results demonstrated that the spin up period of 45 days would be sufficient for the Deep Bay water body. Dry season was simulated on the month of February while wet season was simulated on the month of July. Appendix 5A includes the updated flow and load data within the Study Area.

5.6.4 Scenario Runs

5.6.4.1 Under normal circumstances, the collected raw sewage would be pumped from the Ha Tsuen Pumping Station to San Wai STW for treatment. The treated effluent would be discharged through the NWNT Outfall. The detailed model for water quality computation predicted the impacts due to different treatment options from the San Wai STW. The options for treatment levels include:
Option 1 CEPT
Option 2 CEPT with disinfection
Option 3 Secondary treatment with disinfection
Option 4 Secondary treatment with nitrogen removal and disinfection

5.6.4.2 Comparisons of the model results for these scenarios were made. The difference between the first two options reflected the increase in E. coli levels in the receiving water body. The water quality in terms of BOD levels was the major parameters considered when comparing between Option 2 and Option 3. The effects on the nitrogen levels in the receiving water body were determined by comparison between Option 4 and Option 3.

5.6.4.3 In addition, a scenario run for the case without the project was included to give the baseline conditions. The baseline conditions assumed that the existing treatment level and design capacity at San Wai STW be remained unchanged. Under the baseline scenario, the pollution loads that could not be treated by the existing San Wai STW and the treated effluent from Yuen Long STW will be directed to Deep Bay rather than Urmston Road Outfall.

5.6.4.4 Under the four options, the flow entering San Wai STW is predicted to be less than its design capacity of 264,000m3/s. A sensivitivity test will therefore be conducted to investigate the change in water quality as a result of San Wai STW reaching its full capacity. The level of treatment of the preferred Option concluded in the opertional phase assessment will be adopted in the sensitivity test. Similar to the modelling scenarios, both dry and wet seasons, with a spin up period of 45 days and a simulation period for a 15-day spring-neap cycle, will be performed and the results will be compared with the preferred option.

5.6.4.5 The model results were compared with the relevant WQOs for marine waters of the North Western WCZ to check for compliance.

5.6.5 Emergency Discharges from San Wai STW / Ha Tsuen Pumping Station

5.6.5.1 With reference to Section 1.2 of this report, emergency discharges from San Wai STW / Ha Tsuen Pumping Station would first enter Tin Shui Wai Drainage Channel and eventually be discharged into the Inner Deep Bay waters. The existing water quality in Deep Bay is poor and the pollution in the inner bay is comparatively higher than that in the outer bay. Based on EPD's routine monitoring water quality data, high E. coli and ammonia nitrogen levels and low dissolved oxygen levels were recorded in the Inner Deep Bay indicating that additional pollution load entering the inner bay waters would further deteriorate the existing conditions leading to adverse water quality impacts. Assessment of the potential impacts due to emergency discharges is therefore an important task in this study.

5.6.5.2 Table 5.4 summaries the emergency discharge senarios. As advised by DSD, the longest substantial emergency repairing and maintenance works, though very remote, would be for NWNT tunnel which could have up to 12 days. The worst case of the emergency discharge duration of 12 days would therefore be taken for assessing the very worst scenario. It was further assumed that the discharge would occur at the beginning of flood tide and during the neap tide in the dry and wet seasons. Under these circumstances, the pollutants would be transported to the Inner Deep Bay. Accumulation of pollutants in the Inner Deep Bay may cause adverse water quality impacts. The period of the emergency discharges simulated in the water quality models will be during February (dry season) and July (wet season).

Table 5.4 Emergency Discharge Scenarios

5.6.5.3 In order to assess the magnitude of the water quality impacts, the predicted results from the emergency discharge scenarios were compared with the scenario under normal operation.

5.6.6 Model Output

5.6.6.1 Statistical analysis of water quality changes was conducted at representative indicator points within the study area. Some of the indicator points were located in the same locations as EPD marine water sampling stations and the other indicator points represented the water quality sensitive receivers.

5.6.6.2 The EPD marine water sampling stations include DM1, DM2, DM3, DM4 and DM5 within the Deep Bay WCZ, and NM1, NM2, NM3, NM5, NM6 and NM8 within the North Western WCZ. The predicted mean values for the dry and wet seasons were averaged to represent the annual mean results. Annual mean depth-averaged results in the form of table, contour plot and time series plot for depth-averaged 90%ile DO, bottom 90%ile DO, TIN, SS, UIA and E. coli are presented in Section 5.8.2. The annual averaged data presented in the tables and contour plots are arithmetic means results except for E.coli which are annual geometric means.

5.6.6.3 The contour plots show the spatial distribution of the concerned parameters and cover all the indicator points selected for the assessment. The annual depth-averaged results at all the indicator points for the scenario runs including the case without the project are summarised in tables for comparison.

5.6.6.4 For the discharges via the NWNT Outfall, time series plots for the parameters of depth-averaged UIA, TIN and E. coli at the following indicator points are presented:

· EPD marine water sampling stations NM5 and DM5
· The Marine Park at Sha Chau/Lung Kwu Chau
· Chinese White Dolphin feeding ground in the Urmston Road Channel
· Cooling water intake for CLP Black Point Power Station

5.6.6.5 For the emergency discharges via the Tin Shui Wai Drainage Channel, time series plots for the parameters of depth-averaged UIA, TIN and E. coli at the following indicator points are presented:
· EPD marine water sampling stations DM1, DM2 and DM3
· Pak Nai Site of SSSI
· Mai Po Nature Reserve in the Inner Deep Bay
· Tsim Bei Tsui SSSI

5.7 Identification of Environmental Impacts

5.7.1.1 The potential water quality impacts that may arise during the construction phase of the project include:
· Construction site runoff;
· Sewage generated from workforce; and
· Accidental spillage of chemicals.

5.7.1.2 During the operational phase, the potential water quality impacts will be mainly related to the effluent discharge from the San Wai STW. Key concerns are:
· Discharge of treated effluent through the outfall at Urmston Road for different treatment options; and
· Emergency discharge to the Deep Bay waters or North Western waters.