54

Ji, et al

Effect of wastewater discharge on Jiangsu coastal environment, China Hong Ji†∞, Shunqi Pan‡ and Longxi Han† †College of Environmental Science and Engineering, Hohai University, Nanjing, 210098, China [email protected] [email protected]

‡School of Engineering, Cardiff University, Cardiff, CF24 2AA, UK [email protected]

∞School of Marine Science &

Engineering, Plymouth University, Plymouth, PL4 8AA, UK

www.cerf-jcr.org

ABSTRACT Ji, H., Pan, S. and Han, X., 2013. Effect of wastewater discharge on Jiangsu coastal environment, China In: Conley, D.C., Masselink, G., Russell, P.E. and O’Hare, T.J. (eds.), Proceedings 12th International Coastal Symposium (Plymouth, England), Journal of Coastal Research, Special Issue No. 65, pp. 54-59, ISSN 0749-0208. www.JCRonline.org

To meet the needs of the rapid economic development in the coastal area of Qidong City in Jiangsu Province of China, a new wastewater treatment plant and a reclaimed water plant have been planned. The preliminarily treated wastewater will be discharged to the coastal area in the Yellow Sea near Dayang Port. Due to the complex tidal regime in the area, the impacts of the outfall from the treatment plant can be important for the surrounding marine environment. The present study aims to investigate the impacts of the discharge of the treated wastewater into the Yellow Sea under three operational conditions using the TELEMAC software. The computational domain covers a large coastal area (~2800 km2). TELEMAC2D is used to compute the depth-averaged currents, and the convection and diffusion of the pollutants, namely CODMn and Inorganic Nitrogen, which are treated as the neutral buoyancy tracers, but with temporal decay. The hydrodynamic aspects of the model are calibrated with the field measurements. The model results show a good agreement of the computed water levels and velocities compared with the field measurements. Model applications include the impact studies with the normal operational conditions of both wastewater treatment plant and reclaimed water plant, and two accidental operational conditions with increased discharges of treated and untreated wastewater. The results show that when both plants are operating under normal full capacity conditions, the treated wastewater discharged to the sea poses no threat to the water quality and marine environment under the current regulations. However, when the untreated wastewater is discharged in the sea, it will pose a significant threat to the water quality. ADDITIONAL INDEX WORDS: Marine environment, water quality, environmental impact, TELEMAC, hydrodynamic modelling

INTRODUCTION

Recent years have seen a rapidly development in China’s economy, particularly in the coastal areas where the vast marine resources can be exploited for both recreational and industrial purposes. In Jiangsu Province, one of the areas included in a recently published integrated coastal zone development plan along its east coasts is the Lvsi Port area. A wastewater treatment plant, which will have a design discharge capacity of 150,000 m3/d (day), is one of the major projects in the plan. Alongside the wastewater treatment plant, a reclaimed water plant is also planned in order to reclaim part of the treated wastewater back for low grade uses, such as irrigation of agricultural fields and public parks, so that the discharge of the treated wastewater released to the sea can be significantly reduced. The design capacity of the reclaimed water plant is 60,000 m3/d, which is to serve the newly established industrial zone at Qidong near Dayang Port. As a result, the net discharge of the treated wastewater directly into the Yellow Sea can be reduced to 90,000 m3/d, through the existing sea outfall near Dayang Port, see Figure 1. When such a large scale wastewater treatment plant is to be built, the major concern has been the impact of the treated wastewater discharged to the sea on the marine environment in the coastal area immediately adjacent to the sea outfall. The current regulations in China: the Pollution Control Criteria for Wastewater ____________________ DOI: 10.2112/SI65-010.1 received 07 December 2012; accepted 06 March 2013. © Coastal Education & Research Foundation 2013

Disposal Project in Ocean (GB18486-2001), set strict limits of the pollutant concentrations of the treated wastewater into the coastal waters, In addition, it also requires that if the wastewater is discharged into a region larger than 600 km2, the area of mixing zone at any time should be no more than 3.0 km2. In recent years, a number of studies have been carried out in the areas close the outfall, such as the work of Wu et al. (2012) on evaluating the ecological status of the coastal waters in the East China Sea, and Pei et al. (2009) on nutrient dynamics in the Yangtze River Estuary. However, detailed research of the water quality in regional scale, particularly for the wastewater treatment plant is yet to be carried out. Due to the existence of Xiaomiao tidal channel, and the merging point of the tides from the Yellow Sea in the north and tides from the East China Sea in the south, the tidal currents in this region become very complex and rather difficult to model. In the past few years, computer models have been widely used to evaluate the impacts of human activities on the environment. As the marine environment is concerned, the numerical models used for the marine environment studies include the Environmental Fluid Dynamics Code (EFDC) by Shi et al. (2011) and Wan et al. (2012); MIKE by Suresh et al. (2011) and Remya et al.(2012); and DELFT3D by Kuang et al. (2011) and Ying et al. (2012) at different sites worldwide. The TELEMAC software, which was originally developed by EDF and has been recently made opensource code, is one of the commonly used software for modelling hydrodynamics and pollutant transport in coastal and estuarine

Journal of Coastal Research, Special Issue No. 65, 2013

Effect of wastewater discharge on Jiangsu coastal environment, China

55

MODEL DESCRIPTION

The TELEMAC suite consists of various simulation modules: hydrodynamics, waves, sediment transport and morphodynamics, which can also be coupled with water quality modules. The software basically solves the relevant governing equations using algorithms based on the finite-element or finite-volume methods. In the present study, TELEMAC2D (v6.1), implemented with a user-function to deal with the concentration decay of the tracers is used. The computer model solves the following governing equations for depth-average free surface flow, and transport of dissolved tracers in Cartesian coordinates:

Figure 1. Study area and the locations of the outfall (∆), Lvsi tide gauge () and velocity measurement station #4 (). areas. Recent applications of TELEMAC throughout the Europe include the works from Abadie et al. (2008), Herbert et al. (2011); Monteiro et al. (2011) and Bedri et al. (2011). Due to the availability, TELEMAC2D was adopted and employed in the present study, as the first study of this kind for this region, to model the hydrodynamics and pollutant transport and to assess the impacts of treated wastewater discharge on the surrounding area of the outfall with the normal operational conditions of both wastewater treatment and reclaim water plants, and two accidental operational conditions. Two popular control indexes for water quality, namely CODMn and inorganic nitrogen are considered as the assessment factors. The depth-averaged hydrodynamics and the concentrations of the pollutants are modelled by TELAMAC2D under 3 scenarios, consisting of the normal operational condition and two accidental operational conditions of a wastewater treatment plant and a wastewater reclaimed plant. The distributions of the pollutants and the affected areas are studied, so that water quality in the area can be assessed and high risk areas are identified. It is hoped that the model used in this study can provide the essential and detailed information to support the decision making for the project planning and marine environment risk management.

 h    u  h  hdiv (u )  S h t u   z 1  u  u   g  div (h t u )  S x t x h v   z 1  u  v   g  div (h t v)  S y t y h T   1  u  T  div (h T T )  ST t h

(1) (2) (3) (4)

where, t = time; x and y = longitudinal and latitudinal Cartesian coordinates respectively; h = water depth; u and v = longitudinal and lateral velocities, respectively; T = passive (non-buoyant) tracer; g = gravity acceleration; t and T = diffusion coefficients for flow and tracer respectively; z = free surface elevation; Sx and Sy = source or sink terms for flow; ST = source or sink term for tracer. Unstructured triangular mesh is generated with its preprocessor, as shown in Figure 2, where the mesh becomes progressively finer towards the coastline, where the type size of the mesh is about 10 m. The mesh consists of 29,201 nodes in total.

STUDY AREA

The study area is located at the east coast of Jiangsu Province near Dayang Port (see Figure 1). The computational domain covers a coastal area approximately 2800 km2 along about 27 km coastline, as shown in Figure 2. The mean water depth at the offshore boundary of the computational domain is about 15-18 m. The geographical coordinate of Lvsi Port is 121°40′E, 32°08′N, some 60 km north to the Yangtze River estuary. The Xiaomiao tidal channel lies 5 km away from the shoreline. There are a number of sandbanks within the computational domain, namely Yaosha in the north and Hengsha in the south. The existence of the main channel and sandbanks greatly complicates the tidal flows in the area. The tides in this area can be characterised as semi-diurnal tides, with spring and neap tide ranges being 5 m and 2 m respectively and the average tide range of 3.53m. The tides are slightly asymmetric with the average flood and ebb tide durations of 6 hours 19 minutes and 6 hours 6 minutes respectively. Both the local temperature and salinity are found fairly evenly distributed in the water column except in summer seasons.

Figure 2. Computational grid and bathymetry. The time discretization used in TELEMAC2D is semiimplicit, with a Crank-Nicolson factor of 0.6. The equations are solved using Method of Characteristics, PSI distributive scheme and N distributive scheme for advection for flow velocities, water depth and tracers respectively. As for the PSI scheme, the Courant–Friedrichs–Lewy condition (CFL condition) must be considered for convergence, and commonly the Courant number should be less than 1. Therefore, in this study, the time step is set

Journal of Coastal Research, Special Issue No. 65, 2013

Ji, et al

concentrations are 16.67 mg/L for CODMn and 10 mg/L for Nitrogen. Scenario 2 presents the conditions that the reclaimed water plant stops working, so that all treated wastewater of 150,000 m3/d is released into the sea, with the same pollutant concentrations as those in Scenario 1. Scenario 3 represents the worst case when the reclaimed water plant stops working, and the wastewater treatment plant operates abnormally. This will result in the wastewater of 150,000 m3/d being released directly into the sea untreated, also with much higher pollutant concentrations: 166.67 mg/L for CODMn and 70 mg/L for Nitrogen.

Comp_Z

3.0

RESULTS AND DISCUSSION

Following model validation, three test cases, the conditions of which are listed in Table 1, are carried out. These conditions represent three different operational modes of the wastewater treatment plant and the reclaimed water plant. Scenario 1 presents the normal operational conditions for both plants. Under this scenario, the wastewater treatment plant operates normally, with the full discharge capacity of 150,000 m3/d, whilst the reclaimed water plant also operates normally with its full recycling capacity of 60,000 m3/d. As a result, the treated wastewater released directly into the Yellow Sea is 90,000 m3/d, and the pollutant

2.0 1.0

0.0 -1.0

-2.0 -3.0 -4.0 1000

1010

1020

1030

1040

1050

1060

1070

1080

494

504

514

524

Time (hrs) Comp_Z

Neap tide

Mea_Z

4.0 3.0

MODEL VALIDATION

2.0 1.0

0.0 -1.0

-2.0 -3.0 -4.0 444

454

464

474

484

Time (hrs)

Figure 3. Comparison of the computed and measured tidal elevations at Lvsi Station during spring and neap tides.

2

4

Comp V

Meas U

Meas V

Z

1.5

3

1

2

0.5

1

0

0

-0.5

-1

-1

-2

-1.5

Water Level (m)

Velocity (m/s)

Comp U

-3

-2 1000

1005

1010

1015

1020

1025

-4 1030

Time (s) 1.5

1.5 Comp U

Comp V

Meas U

Meas V

Z

1

0.5

0.5

0

0

-0.5

-0.5

-1

Water Level (m)

1

Velocity (m/s)

TELEMAC2D was validated against the field data (Chen, 2012) over the period from 17:00 26th to 20:00 27th, March, 2009 for typical spring tides and from 9:00 4th to 12:00 5th, March, 2009 for neap tides. The comparisons of the computed tidal elevations and current velocities with the measurements at the Lvsi tidal gauge location and the current measuring location #4 are shown in Figures 3 & 4 respectively, while the locations of the measurements are shown in Figure 1. It can be seen that in general the results show a good agreement between the model predictions and the field data. The tides at the study area are mainly controlled by the progressive wave from the East China Sea, as the tidal orbits show that the main tidal flow direction is aligned with the axis of the Xiaohongmiao Channel (not shown here). Close examination of the tidal levels shown in Figure 3 indicates that the computed water levels agree better during the flood phase than those during the ebb phase for spring tides. For neap tides, the agreement is generally better for both phases. As shown in Figure 4, the agreement between the computed tidal velocities with the measurements is generally good, but the results indicate that for spring tides, there is a small discrepancy in current direction, as the V component of the flow velocity is slightly under-predicted. However, for neap tides, the agreement for both the velocity magnitude and phase is satisfactory. The results, nevertheless, show that the model has been set up correctly, and the parameters are adequately used. More importantly, the tidal elevations imposed along the open boundary, which was done through a user function, have worked properly and satisfactorily.

Spring tide

Mea_Z

4.0

Tidal Level (m)

to 60s throughout the simulations for all cases to ensure the stability of the computations. TELEMAC2D provides a variety of options for bottom friction. In this study, Stricker’s Law which describes energy dissipation is used by assuming a constant friction coefficient throughout the domain. Sensitivity tests for the bottom friction were carried out, but its effect was found to be insignificant. Therefore, the friction coefficient is set to 50. For the simplicity, a constant viscosity of 10 m2/s suggested by the software was used. Along the offshore open boundary, tidal elevations are required to be specified in the model. In order to better represent the complex tidal currents in the area, a user function was implemented in TELEMAC2D, so that the tidal elevation at each node point at the open boundary can be specified individually to allow the temporal and spatial variations. The tidal elevations were calculated using OTPS (Organ State University Tidal Prediction Software), with the bathymetry resolution of 1/30° in the area. The tidal output includes 8 tidal constituents, namely: M2, S2, N2, K2, K1, O1, P1 and Q1. The computer simulations were carried out over 90 tidal cycles from 00:00:00 13th February to 00:00:00 30th March 2009. Both pollutants (CODMn and Nitrogen) are treated as neutral buoyancy tracers, but with temporal decay included in the model.

Tidal Level (m)

56

-1

-1.5

-1.5 466

471

476

Time (s)

481

486

491

496

Figure 4. Comparison of the computed and measured current velocities (U & V) at Station 4# during spring and neap tides (Z = water level). Table 1. Operational scenarios Scenario Discharge (t/d) 1 90,000

CODMn (mg/L) 16.67

Nitrogen (mg/L) 10.00

2

150,000

16.67

10.00

3

150,000

166.67

70.00

The computation for each scenario is carried out for 46 days (=1103 hours) which covers three spring-neap tidal cycles. The computed concentrations of CODMn and Nitrogen during a spring

Journal of Coastal Research, Special Issue No. 65, 2013

Effect of wastewater discharge on Jiangsu coastal environment, China

tide, a neap tide and the maximum over the 46-day computational period are shown in Figures 5 – 7. Figure 5 shows concentrations of CODMn and Nitrogen at the high water levels during a spring tide (left panel - at 661 hours), a neap tide (middle panel - at 864 hours) and the maximum concentration (right panel). It can be seen that in this case (Scenario 1) the spreading of the pollutants is restricted in a small area around the outfall. The concentrations are very low during spring tides as expected. During neap tides, the concentrations are noticeably higher due to the weaker diluting power from the tides. The concentrations of both pollutants are found lower than 0.05 mg/L. The maximum concentrations of both pollutants over the computational duration are in a similar range. Taking into consideration of the background concentration of 0.255 mg/L for neap tides as listed in Table 2, and assuming a linear accumulation, the maximum concentration for CODMn is approximately 0.3 mg/L, which is below the regulatory limit of 4.0 mg/L, see Table 2. For Nitrogen, the maximum concentration is found to be approximately 0.123 mg/L. Combining this concentration with the background concentration of 0.230 mg/L for neap tides gives a resulting concentration of 0.353 mg/L, which is also below the regulatory limit of 0.4 mg/L. Therefore, for the normal operational conditions of both wastewater treatment and reclaimed water plants, the water quality in the study area is satisfactory. Table 2. Background and regulatory concentration limits CODMn Nitrogen Concentration (mg/L) Background

Spring tides

0.194

0.209

Neap tides

0.255 4.000

0.230 0.400

Regulatory limits

Spring Tides

57

For Scenario 2, the computed results exhibit the similar patterns to those in Scenario 1. The maximum concentrations of CODMn and Nitrogen over the computational duration are shown in Figure 6. It can be noticed that the spreading of the pollutants is wider. The maximum concentrations for both pollutants are higher than those in Scenario 1. The maximum concentration of CODMn is found to be around 0.336 mg/L. Combining this with the background concentration gives the resulting concentration of CODMn 0.591 mg/L. This again is within the regulatory limit. However, the maximum concentration for Nitrogen is found to be 0.204mg/L, almost twice as high as that in Scenario 1. Together with the background concentration for neap tides, the resulting concentration is 0.434 mg/L, and exceeds the regulatory limit of 0.4 mg/L. But it should be pointed out that high concentration of Nitrogen is only limited in a small area. As the receiving area of the treated wastewater is larger than 600 km2, it is within the regulatory requirements if the area with high concentration is no larger than 3.0 km2 at any time. For Scenario 3, the computed concentrations of both pollutants are shown in Figure 7, similar to those in Figure 5. In comparison, it can be found that there is a large area where the concentrations of both pollutants are higher than 0.15 mg/L, the upper scale of the figure set for inter-comparison reasons. It is also important to notice that the high concentration area is mainly close to the shoreline, which can impose a greater threat to the public health. The maximum concentration of CODMn is 3.36 mg/L which is just below the regulatory limit. However, the maximum concentration for Nitrogen is 1.429 mg/L, which is far exceeding the regulatory limit. The affected area is also much more extended during the neap tides. In order to examine further the affected area in Scenario 3, the total areas where the concentration of each pollutant is higher than the regulatory limit over the computational duration are shown in

Neap Tides

Maximum

Figure 5. Computed concentrations of CODMn (top) and Nitrogen (bottom) for Scenario 1 during a spring tide (left), neap tide (middle) and maximum over the computational duration (right). Journal of Coastal Research, Special Issue No. 65, 2013

58

Ji, et al

Figure 6. Computed maximum concentrations over the computational duration: CODMn (left) and Nitrogen (right) for Scenario 2

Spring Tides

Neap Tides

Maximum

Figure 7. Computed concentrations of CODMn (top) and Nitrogen (bottom) for Scenario 3 during a spring tide (left), neap tide (middle) and maximum over the computational duration (right). Figure 8. The results show that for CODMn, the concentration is generally low, so there is no area where its concentration exceeding the regulatory limit. For Nitrogen, within the first two spring-neap tidal cycles, the affected area is markedly large during neap tides, but is less than 30 km2. However, after the third neap tides, the area affected by the higher concentration Nitrogen is considerably increased. Close examination of the evolution of the

affected area through the tidal cycles (not shown here) finds that the concentration is built up gradually with more peaky distribution in a confined area, but it reaches the peak before the third neap tides, and then the concentration of Nitrogen disperses rapidly to the surrounding area, which results in the large area being over the regulatory limits.

Journal of Coastal Research, Special Issue No. 65, 2013

Effect of wastewater discharge on Jiangsu coastal environment, China

59

Figure 8. The total area affected by the pollutants when the concentrations exceeding the regulatory limits taking account for the background concentrations during spring tides (L – low) and neap tides (H –high) for Scenario 3.

CONCLUDING REMARKS

TELEMAC2D was set up at the Lvsi Port area in Jiangsu Province at the east coast of China to study the impacts of the treated wastewater discharged into the Yellow Sea on the coastal water quality and marine environment in the area. Concentrations of CODMn and Nitrogen are modelled under the tidal conditions with three scenarios pertaining to different operational conditions of a wastewater treatment plant and a reclaimed water plant. The results show that when both plants are operating under normal full capacity conditions, the treated wastewater discharged into the sea poses no threat to the water quality and marine environment under the current regulations. However, when the untreated wastewater is discharged into the sea, it will pose a significant threat to the water quality. The concentration of Nitrogen in particular will be much higher than the regulatory limits, and the affected area is much larger than that is allowed. Therefore, appropriate mitigation measures must be put in place to reduce the risk of environmental disaster in the area should those conditions occur. The model has provided useful information to help assess the coastal environment and ecological system, but the work can be improved by, for example, including the wave effects and 3-dimensionality.

ACKNOWLEDGEMENTS

The authors would like to thank Professor Yongping Chen and Dr. Guoxian Huang for kindly providing the field measurement data and bathymetry. Thanks also go to the Chinese Scholarship Council for the financial support and Plymouth University for hosting the visiting research student (HJ).

and vertical mixing in Hong Kong waters. Journal of Coastal Research, 27(6A), 74-86. Monteiro, I.O., Marques ,W.C., Fernandes ,E.H., Gongalves, R.C. and Moller Jr, O.O., 2011. On the effect of earth rotation, river discharge, tidal oscillations, and wind in the dynamics of the Patos Lagoon Coastal Plume. Journal of Coastal Research, 27(1), 120-130. Pei, S.F., Shen, Z.L and Laws, E.A., 2009. Nutrient dynamics in the upwelling area of Changjiang (Yangtze River) estuary. Journal of Coastal Research, 25(3), 569-580. Remya, P.G., Kumar R.A.J., Basu, S. and Sarkar, A. 2012. Wave hindcast experiments in the Indian Ocean using MIKE 21 SW model, Journal of Earth System Science, 121(2), 385-392. Shi, J.H., Li, G.X. and Wang, P., 2011. Anthropogenic Influences on the tidal prism and water exchanges in Jiaozhou Bay, Qingdao, China. Journal of Coastal Research, 27(1), 57-72. Suresh, P.K.; Sundar, V. and Selvaraja, A, 2011. Numerical modeling and measurement of sediment transport and beach profile changes along southwest coast of India. Journal of Coastal Research, 27(1), 26-34. Wan, Y.S, Ji, Z.G., Shen, J.Hu, G.D.and Sun, D.T., 2012. Three dimensional water quality modeling of a shallow subtropical estuary. Marine Environmental Research, 82, 76-86. Wu, H.Y., Chen, K.L., Chen, Z.H., Chen, Q.H., Qiu, Y.P., Wu, J.C. and Zhang, J.F., 2012. Evaluation for the ecological quality status of coastal waters in East China Sea using fuzzy integrated assessment method. Marine Pollution Bulletin, 64(3), 546-555. Ying, X.M., Ding, P.X., Wang, Z.B. and Dirk, S.V., 2012. Morphological impact of the construction of an offshore Yangshan Deepwater Harbor in the Port of Shanghai, China. Journal of Coastal Research, 28(1A), 163-173.

LITERATURE CITED

Abadie, S.M., Briere, C, Dubranna, J., Maron,P. and Rihouey, D.,2008. Erosion generated by wave-Induced currents in the vicinity of a jetty: case study of the relationship between the Adour River Mouth and Anglet Beach, France. Journal of Coastal Research, 24(1), 59-69. Bedri, Z., Bruen, M., Dowley, A. and Masterson, B., 2011. A three-dimensional hydro-environmental model of Dublin Bay. Environmental Modeling and Assessment, 16(4), 369-384. Chen, Y., 2012. Field measurements at Jiangsu East Coasts, Personal communication. Herbert R.J.H., Willis, J., Jones, E., Ross, K., Hubner, R., John, Humphreys, J., Jensen, A. and Baugh, J., 2011. Invasion in tidal zones on complex coastlines: modelling larvae of the non-native Manila clam, Ruditapes philippinarum, in the UK. Journal of Biogeography, 39(3), 585-599. Kuang, C.P., Lee, J.H.W., Harrison, P. J. and Yin, K.D., 2011. Effect of wind speed and direction on summer tidal circulation

Journal of Coastal Research, Special Issue No. 65, 2013