Contamination of Bourgas Port Waters with Oil

Transcription

1 Contamination of Bourgas Port Waters with Oil Vasko Galabov (1), Anna Kortcheva (1,2), Georgi Kortchev (1,3) and Jordan Marinski (1,4) (1) National Institute of Meteorology and Hydrology, Bulgarian Academy of Sciences, 1784 Sofia, Bulgaria Tel: Vasko.Galabov@meteo.bg (2) Tel: Anna.Kortcheva@ meteo.bg (3) Tel: Georgi.Kortchev@ meteo.bg (4) Tel: marinski@bas.bg Abstract Accurate prediction of the behavior of oil pollution is crucial for successful decision making and planning of operations to combat pollution of the coastal zone to protect the environment. In addition, the oil spills in the Bourgas Bay of West Black Sea may cause extra contamination of port waters in different terminals of Port of Bourgas with oil and liquid dangerous products which is extremely undesirable from the environmental point of view. The present study aims to assess the impact of the risk of oil spills in the auquatory of Bourgas Port and to determine potentially dangerous areas of contamination in the inner basins of the port and the conditions for their occurrence. The numerical model MOTHY, developed by Météo-France, have been used to predict the drift of pollutants on the sea surface. Numerical simulations are carried out in the Bourgas bay and harbor waters for potential and real oil spill accidents. The range of hypothetical scenarios includes various winds and current conditions. The simulation results indicate the places with a high risk of oil pollution in the port of Bourgas. The proposed methodology utilized in this study is applicable to risk assessments for other ports and coastal areas potentially affected by floating pollutants.

2 Introduction Port of Bourgas is located in the Bourgas Bay that is the widest bay in the western part of the Black Sea. Set deep into the mainland, the bay includes the westernmost point (27º26 54 ) of the Black Sea. The Bourgas Bay is 41 km at its widest part and 25 m at its deepest, reaching 31 km at its greatest innermost extent, and 750 km 2 area, approximately. The Port Bourgas is the biggest port in Bulgaria and it is a key-point of Trans border Corridor 8, ensures direct access to the Black Sea of Southern Italian ports. For this reason the Port of Bourgas is chosen as a pilot site in two South East European projects ECOPORT 8 and TEN ECOPORT ( and the present study is a part of them. According to BG Ministry of transport, Bourgas port,is divided the following terminals: East, Bulk cargoes, Bulk terminal 2A, West, Oil terminal Rossenetz and passenger terminals in Nesebar and Sozopol (Fig. 1). Port activities /in public and private terminals and ports/ interact with activities of adjacent industrial zones around Bourgas Bay and current environment conditions of Bourgas Port are very complicated. The crude oil and oil products terminal Rossenetz has been built about 3 miles southeast of the existing East and West Terminals and its activities are supposed to be one of the main sources for contamination of port waters (internal basins 1, 2, 3 and 4) with oil and oil products. Fig. 1: Location of Bourgas Port terminals

3 It is supposed that the oil spills in the Bourgas Bay may cause extra contamination of port waters in different terminals of the Bourgas port with oil and liquid dangerous products, which is extremely undesirable from an environmental point of view and with respect to the forthcoming certification of the Bourgas port. A powerful tool for determining any potentially dangerous areas of contamination in the inner basins of the port and its terminals is to use the numerical modeling and simulation of the behavior of oil spills in coastal waters. Oil spills in the Bay of Bourgas can be caused by the tankers that deliver the product to the oil terminal Rossenetz of the Bourgas Port supplying LUKOIL refinery with crude oil. To minimize the damage resulting from oil spill accidents, it is necessary to respond quickly with an appropriate strategy. Accurately forecasting the behavior of the spilled oil is crucial for a successful recovery operation and protection of the marine resources. It is also important to simulate the most probable scenarios (i.e. locations on the tanker routes and different weather conditions and circulation within the bay) in order to have a better preparedness and to choose the right positions of the monitoring facilities. Therefore, the numerical modeling and simulation of the spilled oil at sea plays an important role in the oil spill response and contingency decision-making. The study of the effects of the potential oil pollution for the port of Bourgas terminals includes determination of the potentially affected areas of the port waters. Model description Lagrangian trajectory modeling: floating pollutants spread model The oil drift forecast is carried out by means of a numerical model, named MOTHY, describing the ocean dynamics and the physicochemical behavior of the pollutant. The numerical model MOTHY has been developed by the Marine Forecast Division of Meteo-France and implemented for the Black Sea. (Daniel, et al, 1999, 2005, Kortchev et al, 1999). The model MOTHY uses the Langragian approach, trajectories based on the assumption that the oil can be idealized by a large number of particles that move by advection, turbulent dispersion and oil spreading dynamics. Barotropic model MOTHY model system consists of a model for the wind currents and a model for the oil spill drift prediction. The wind currents are computed by a barotropic model coupled with a turbulent model and driven by the wind and mean sea level pressure output from an atmospheric model (global scale or high resolution regional scale). The barotropic model is two-dimensional, coupled with a one-dimensional for the vertical structure of the currents, therefore the wind currents model is 2.5 dimensional depth-integrated and solves the shallow water equations on a regular grid: q 1 q. q f. kq g. Pa ( s b) A. t. H 2 q (1)

4 ( H. q) 0 t (2) where t is the time, q- depth integrated current, η is the sea surface elevation, H- the total depth, f- the Coriolis parameter, k- the unit vector in the vertical, P a is the mean sea level pressure, s is the surface wind stress, b is the bottom friction stress, is the water density, g the gravitational acceleration and A the horizontal diffusion coefficient. These equations are solved by a forward in time integration in Arakawa C-grid, using a finite split-explicit difference scheme. The surface wind stress is calculated by using a quadratic relation (3):. C. W. W sx a d 10 10x. C. W. W sy a d y (3) where W, W are the horizontal components of wind at 10 m above the sea surface, a 10x 10y is the air density and C d is the drag coefficient, calculated by using the relation (4): (4) The bottom stress is calculated as: ( U V ) 1/ U (5) bx C b... V (6) by 2 2 1/ 2 C b.( U ) Here C b is the bottom friction coefficient (the value is in our case). The boundary conditions are: at the bottom the normal component of the current is 0 and at the surface condition for the gravitational waves. Oil spill drift model The second component of the system is the oil spill drift model. The oil spill is modeled as an ensemble of independent droplets, moving in response to shear current, turbulence and buoyancy. The shear current is calculated for each droplet by a bilinear eddy viscosity model. The vertical eddy viscosity in the model increases linearly with the distance from the sea surface and the bottom (Poon and Madsen, 1991). The governing equation is: w 1 P i f w t t n z w z (7)

5 where w=u+i.v is the horizontal current velocity, t is the eddy viscosity. The turbulent diffusion is represented by three-dimensional random walk technique. For one time step t, the motion is: D R 2 K t h h in direction: 2 R (8) where K h is the horizontal diffusion coefficient and R- random number between 0 and 1 In the vertical, the motion is: D R v Kv t (9) where K v is the vertical diffusion coefficient. The buoyancy depends on the density and size of the droplet, so the larger ones are more buoyant and tend to remain on the surface, while the smaller are mixed downward. U f The vertical speed U f is determined by (10) and (11) (Elliot et al, 1986): 2 0 g d ( 1 ) For small droplets d d c (10) Uf gd ( 1 ) 3 For large droplets d d c (11) where the critical diameter d c is: d c 3 9, g( 1 ) is the sea water density and viscosity. If a droplet is moved on the land, then it is considered beached and takes no further part of the simulation. (12) Water circulation in the Bourgas Bay geophysical hydrodynamic model The water circulation in the Bourgas Bay and character of sea currents is determined by geophysical hydrodynamic non-stationary 3D numerical model created in the Institute of Numerical Mathematics (Sarkisyan A., 1977) and developed in the Institute of Oceanology, Varna (Trukhchev et al., 1995, 2004 ). The initial system includes equations and boundary conditions, which are traditional for a model of ocean hydrothermodynamics; their finite-difference approximation is based on the balance of properties within the "boxes" on Arakawa B-grid. The modeling computations of Bourgas

6 Bay currents represent the character of circulation in the bay. The results show the presence of a developed eddy structure of the currents and the existence of rapid adjustment (of several hours) of the movements to the changes in the field of wind. As a rule, in the Small Bourgas bay there forms one main vortex: during north, northeast and east winds the system of currents is cyclonal, and during west, southwest and south winds anticyclonal, Fig 2. Fig.2: Cyclonal (a) and anti-cyclonal (b) type of circulation [cm/s] in Bourgas Bay The cyclonal type of movements is usually dominating. In the water areas immediately close to the coast, coastal jets and small scale eddies are formed. In our study the monthly mean climatic fields of the sea current in the Bourgas Bay are calculated because the anti-cyclonical appears mainly when the wind is from the southwestern quadrant and it such situation the winds and currents are typically much weaker, therefore this case is not of much significance for our study goals. Two-dimensional simulation of oil spills and localization of oil pollution in port This study summarizes the results of the hypothetical oil spill simulations in the Bourgas Bay, the objective of which is to localize oil pollution in internal port water basins 1,2,3 and 4 and to estimate the risk of oil-spill contact to the shoreline in the port terminals and harbor structures, (breakwaters and ship berth quay walls). Forecasts strongly depend on the initial conditions, and therefore on the position of the slicks. Based on the identified spill scenarios and type of oil, oil drift modeling can be done using any combination of release rates and durations. In order to capture the wind and ocean current variations, typically simulations are performed using historical wind conditions and monthly averaged permanent currents direction and speed. The study of trajectories (oil drift) consisted of simulations of a large number of scenarios of combination of incident type, current field and local winds in the Bourgas bay. The oil spill

7 model MOTHY calculates the spreading and diffusion and slick drift of an oil particle at the water surface due to the winds, waves and currents. Scenarios covered spills of the oil type Bunker fuel (Fuel oil No. 6); the oil density was chosen to be 930 kg/m3; the duration of the release 4 hours, the durations of the spill simulations up to 168 hours (7 days). Numerical simulations are carried out in Bourgas Bay for potential and real oil accidents. The range of hypothetical scenarios includes various wind and current conditions. The points of release for the simulations are chosen along the main tanker routes (fairway) to the terminals of Port Bourgas. Also taken into account is the possibility of pollution due to a release of fuel from other type of ship. The numerical simulations were carried out for the area of the Bourgas bay and port waters for 20 locations imitating potential and real oil accidents along the main (existing) traffic route (corridor) to the port of Bourgas (terminal East, terminal 2A, terminal West ) and near to the Oil terminal Rosenetz. The range of the hypothetical scenarios includes various wind and current conditions. The starting points for the simulations are chosen along the main transport route (fairway) to the terminals of the port of Bourgas. Input data for MOTHY model are bathymetry, monthly mean climatic fields of the sea currents in the Bourgas bay calculated by the hydrodynamic non-stationary 3D numerical model, wind and mean sea level pressure fields. The simulations yielded the following area oiling results: In case of an accident in the area of roadsteads or in the area of the approach channel (external port waters) the simulations show that there is a risk for oil pollution mainly for the internal basin 2, fig 3. In case of oil pollution simulation with sources in small Bourgas bay the simulation output shows that the internal basins 2, 3 and 4 are affected by significant oil pollution in various meteorological conditions. The basin 1 of the terminal East (the old harbor of Bourgas) is not affected by oil pollution. A pollution of this terminal occurs only in a case of accidents inside the harbor. In case of an accident east of the roadsteads (east of the oil terminal Rosenetz) the probability for pollution of the terminals in the Bourgas port is much lower and is negligible. Because the main route of the tanker ships is east of Rosenetz (where there is some chance of a big oil accident) the terminals and port waters in basins 1,2,3 and 4 are well protected against such accidents. Taking into account analyses of all results, obtained by processing a large number of simulations we can summarize the results and state locations and areas with different level of oil risk pollution in port waters, Fig. 4.

8 Fig.3: Oil pollution in port waters when oil drift source is located on access channel - locations of oil pollution Fig.4: The zones in port at risk of oil pollution: - area with significant risk will be affected in any conditions; - area with average risk depending on predominant conditions (winds and currents); - area with not negligible risk during the specific meteorological conditions;

9 Conclusions The modeling results present a range of useful information on potential oil movement and behavior in Bourgas Bay and Port of Bourgas. It should be remembered that these results are based on probabilities and that the exact areas, oiled and travel times and locations cannot be predicted with complete accuracy. Modeling can be a very useful response tool but should be used in conjunction with all other available information. Acknowledgements This research was made possible thanks to a research grant provided by South East Europe Transnational Cooperation Program within Transnational enhancement of ECOPORT 8 network - TEN ECOPORT project Code SEE/D/189/2.2/X References Daniel P. (1999). Marine pollution transport modeling. In: Proceedings of Marpolser 98, WMO/TD No960, Daniel P., Josse P.and Danadin P. (2005), Further improvement of drift forecast at sea based on operational oceanography systems. In:Coastal Engineering VII, Modelling, Measurements, Engineering and Management of Seas and Coastal Regions, Elliot A.J., Hurford N. and Penn C.J. (1986), Shear Diffusion and the Spreading of Oil Slicks Marine Pollution Bulletin, Pergamon Journals Ltd., 17, (7), Kortchev G., Mungov G., Kortcheva A.and P. Daniel (1999), Operational Forecasting of oil spill drift in the Black Sea, In: Proc. Oceanogpraphy of the Eastern Mediterranean and Black Sea, Zappeion International Conference, Atthens, Greece, 26 February, 1999, Poon Y.K. and O.S. Madsen, 1991: A two layers wind driven coastal circulation model, J. Geophys. Res., 96, (C2), Sarkisyan A. (1977), The Diagnostic Calculation of a Large Scale Oceanic Circulation. In: The Sea, Marine Modeling. New York, 6, Trukhchev D., Ivanov D., Ibrayev R., Patzireva T. and Rabie A. (2004), Hydrophysical Study of Bourgas Bay. Modelling the Synoptic Circulation Patterns. Comptes Rendus de l'academie Bulgare des Sciences, 57, (3), Keywords: marine pollution, oil spill, numerical simulations, port waters, Black Sea