11 Applications of an Oil Drift and Fate Model for Fairway Design
379
of several kilometers. The vertical transport assumes hydrostatic balance and incompressibility of sea water. Horizontal transport is modelled using the Boussinesq approximation, where density differences are neglected in all but gravity terms.
Higher order contributions to the dynamics are parameterized following Smagorinsky (1963) in the horizontal direction, and using a k–ω turbulence closure, which
has been extended for buoyancy-affected geophysical flows (Umlauf et al. 2003) in
the vertical direction. The turbulence model includes a parameterization of breaking
surface and internal waves. Stability functions from Canuto et al. (2001, 2002) for
the vertical eddy diffusivities of salinity, temperature and momentum have been applied. The numerical implementation uses a staggered Arakawa C-grid and z-level
coordinates, a flux-corrected horizontal advection scheme and free-slip conditions
along the coastlines.
Boundary conditions at the sea surface and sea bed use the quadratic dependence
of the wind induced shear stress and frictional bottom stress on the wind speed
and current speed, respectively. Following Andrejev et al. (2004) the wind drag
coefficient was set to C D = (0.63 + 0.066U 10 ) × 10 −3 , where U 10 is the wind speed
at a standard 10 m height. Unrealistic strong wind forcing during storms is avoided
by making the wind stress parameterization dependent on the effective wind velocity
in relation to the current velocity. At the sea bed, a constant friction coefficient R =
0.0021 has been applied. The model is two-way coupled with a Hibler-type sea ice
model (Hibler 1979) that handles both ice dynamics and thermodynamics (Kleine
1994; Dick et al. 2001). To account for the penetration of short-wave radiation into
the subsurface layers a proper parameterization of the insolation suited for the Baltic
Sea (Meier 2001) is implemented.
The DMI’s operational model for oil drift and fate simulation has been originally developed by the BSH (Dick and Soetje 1990) for the German Bight. The
model was successfully applied to the North Sea and Baltic Sea by the DMI and
BSH, and to West Greenland waters by the DMI. This model has been developed
for tide-dominated seas that occasionally host very rough wave conditions such
as the North Sea, with shallow near coastal waters and limited impact of sea ice
on oil drift and fate. The model has still been successful in the Baltic Sea conditions. The presence of ice cover in a part of this basin during a substantial fraction of the year is accounted for through dynamics of the sea that depends on sea
ice.
The model is able to deal with ship driving fuel oils and a variety of natural crude
oils. The classes of crude oils include extra heavy and heavy crude oils (IFO 450,
Crude Venezuela) with high solidification points and low evaporation, heavy fuel
oils (Bunker C), conventional crude oils (Crude Statfjord, Crude Ekofisk, Crude
Nigeria), light crude oils (Crude No. 2) and light volatile petroleum products with
high evaporation (such as kerosene). Each oil class is composed of 8 fractions with
different boiling points and molecular weight. For the Gulf of Finland simulations
conventional crude oils (Crude Statfjord) were selected.
As discussed above, oil in sea water undergoes several transformations. The
model considers atmospheric interaction processes (evaporation), wave-driven
ocean surface processes (water-in-oil emulsification), horizontal dispersion and
379
of several kilometers. The vertical transport assumes hydrostatic balance and incompressibility of sea water. Horizontal transport is modelled using the Boussinesq approximation, where density differences are neglected in all but gravity terms.
Higher order contributions to the dynamics are parameterized following Smagorinsky (1963) in the horizontal direction, and using a k–ω turbulence closure, which
has been extended for buoyancy-affected geophysical flows (Umlauf et al. 2003) in
the vertical direction. The turbulence model includes a parameterization of breaking
surface and internal waves. Stability functions from Canuto et al. (2001, 2002) for
the vertical eddy diffusivities of salinity, temperature and momentum have been applied. The numerical implementation uses a staggered Arakawa C-grid and z-level
coordinates, a flux-corrected horizontal advection scheme and free-slip conditions
along the coastlines.
Boundary conditions at the sea surface and sea bed use the quadratic dependence
of the wind induced shear stress and frictional bottom stress on the wind speed
and current speed, respectively. Following Andrejev et al. (2004) the wind drag
coefficient was set to C D = (0.63 + 0.066U 10 ) × 10 −3 , where U 10 is the wind speed
at a standard 10 m height. Unrealistic strong wind forcing during storms is avoided
by making the wind stress parameterization dependent on the effective wind velocity
in relation to the current velocity. At the sea bed, a constant friction coefficient R =
0.0021 has been applied. The model is two-way coupled with a Hibler-type sea ice
model (Hibler 1979) that handles both ice dynamics and thermodynamics (Kleine
1994; Dick et al. 2001). To account for the penetration of short-wave radiation into
the subsurface layers a proper parameterization of the insolation suited for the Baltic
Sea (Meier 2001) is implemented.
The DMI’s operational model for oil drift and fate simulation has been originally developed by the BSH (Dick and Soetje 1990) for the German Bight. The
model was successfully applied to the North Sea and Baltic Sea by the DMI and
BSH, and to West Greenland waters by the DMI. This model has been developed
for tide-dominated seas that occasionally host very rough wave conditions such
as the North Sea, with shallow near coastal waters and limited impact of sea ice
on oil drift and fate. The model has still been successful in the Baltic Sea conditions. The presence of ice cover in a part of this basin during a substantial fraction of the year is accounted for through dynamics of the sea that depends on sea
ice.
The model is able to deal with ship driving fuel oils and a variety of natural crude
oils. The classes of crude oils include extra heavy and heavy crude oils (IFO 450,
Crude Venezuela) with high solidification points and low evaporation, heavy fuel
oils (Bunker C), conventional crude oils (Crude Statfjord, Crude Ekofisk, Crude
Nigeria), light crude oils (Crude No. 2) and light volatile petroleum products with
high evaporation (such as kerosene). Each oil class is composed of 8 fractions with
different boiling points and molecular weight. For the Gulf of Finland simulations
conventional crude oils (Crude Statfjord) were selected.
As discussed above, oil in sea water undergoes several transformations. The
model considers atmospheric interaction processes (evaporation), wave-driven
ocean surface processes (water-in-oil emulsification), horizontal dispersion and