11 Applications of an Oil Drift and Fate Model for Fairway Design
377
The oil slick thickness decides which spreading phase is active. The last two phases
depend strongly on oil viscosity and thus on the oil composition and ambient temperature. The ability of oil to form a quasi-stable emulsion with sea water, with
viscosity increase by several orders of magnitude (Børresen 1993), plays a key role
in these processes. One cubic metre of oil can be spread over 0.1 km 2 , so that the
terminal slick thickness is just a few tens of microns, depending on the outcome of
the emulsification process.
Vertical spreading is governed by buoyancy forces and wave-induced mixing:
downward mixing occurs due to turbulence generated by surface waves in the upper layer whereas buoyancy forces may resurface the oil. The mixing depth mainly
depends on the wave height and wind speed. When mixed down, the surface film is
more or less evenly distributed over the mixing depth. This results in a reduction of
its toxic level (in units of oil volume per water volume) to some extent. For example, if an oil slick of thickness 10 µm with zero water content (corresponding to a
volume of 10 m 3 /km 2 ) is mixed down to 10 m depth, the resulting concentration is
1,000 ppb (part per billion) total hydrocarbon.
Similar ensemble particle tracking methods, with a chemico-physical superstructure built upon a basic advection scheme, are used in the modelling of atmospheric
pollution and in simulations of three-dimensional spreading of, e.g., radioactive substances or volcanic ash (Sørensen 1998; Sørensen et al. 2007). This approach has
proved valuable in several incidents with large-scale impact and duration of weeks
to months such as Chernobyl in 1986 (Brandt et al. 1998, 2002) and the Fukushima
accidents in 2011 (e.g., Dietze and Kriest 2012, among many others), or the Eyjafjallajökull eruption in 2010 (e.g., Webster et al. 2012).
In real-time applications, a large part of the information required to carry out an
oil spill simulation is often based on rough estimates. The amount of spilled oil or
the rate of its release are largely unknown. The ambient sea water density and temperature at the accident site may vary in quite a large range but their characteristics
at the accident site and along the pathways of the spill are frequently estimated from
the relevant climatological values. Sea ice conditions may also not be known with
great precision and wave conditions are often parameterized using some simple relationships. Trajectories may still be calculated with a high degree of accuracy, but the
description of the chemical and physical evolution of the spill, which is important
for oil combating, will suffer as a consequence.
Oil spill modelling in ice covered water is not widely described in the literature
(Reed and Aamo 1994) although there exists a certain pool of experiments (Yapa
and Belaskas 1993) and observations (Wang et al. 2008). The wind drift is reduced
in the presence of sea ice depending on the ice concentration (Leppäranta 2010).
The process is discussed in Johansen et al. (2005), Drozdowski et al. (2011). Sea ice
influences oil drift in dense ice (ice concentration larger than 70 %) while there is no
influence in low ice concentrations (less than 30 %). For oil at the surface, the drift
velocity is a combination of current speed and ice drift velocity, with proper weights
assigned depending on the ice concentration. This is a rather crude approach, limiting the effect of the ice to the drift velocity. weathering processes are significantly
slower in ice leads, but the spatial scale is of the order of meters, which is typically
377
The oil slick thickness decides which spreading phase is active. The last two phases
depend strongly on oil viscosity and thus on the oil composition and ambient temperature. The ability of oil to form a quasi-stable emulsion with sea water, with
viscosity increase by several orders of magnitude (Børresen 1993), plays a key role
in these processes. One cubic metre of oil can be spread over 0.1 km 2 , so that the
terminal slick thickness is just a few tens of microns, depending on the outcome of
the emulsification process.
Vertical spreading is governed by buoyancy forces and wave-induced mixing:
downward mixing occurs due to turbulence generated by surface waves in the upper layer whereas buoyancy forces may resurface the oil. The mixing depth mainly
depends on the wave height and wind speed. When mixed down, the surface film is
more or less evenly distributed over the mixing depth. This results in a reduction of
its toxic level (in units of oil volume per water volume) to some extent. For example, if an oil slick of thickness 10 µm with zero water content (corresponding to a
volume of 10 m 3 /km 2 ) is mixed down to 10 m depth, the resulting concentration is
1,000 ppb (part per billion) total hydrocarbon.
Similar ensemble particle tracking methods, with a chemico-physical superstructure built upon a basic advection scheme, are used in the modelling of atmospheric
pollution and in simulations of three-dimensional spreading of, e.g., radioactive substances or volcanic ash (Sørensen 1998; Sørensen et al. 2007). This approach has
proved valuable in several incidents with large-scale impact and duration of weeks
to months such as Chernobyl in 1986 (Brandt et al. 1998, 2002) and the Fukushima
accidents in 2011 (e.g., Dietze and Kriest 2012, among many others), or the Eyjafjallajökull eruption in 2010 (e.g., Webster et al. 2012).
In real-time applications, a large part of the information required to carry out an
oil spill simulation is often based on rough estimates. The amount of spilled oil or
the rate of its release are largely unknown. The ambient sea water density and temperature at the accident site may vary in quite a large range but their characteristics
at the accident site and along the pathways of the spill are frequently estimated from
the relevant climatological values. Sea ice conditions may also not be known with
great precision and wave conditions are often parameterized using some simple relationships. Trajectories may still be calculated with a high degree of accuracy, but the
description of the chemical and physical evolution of the spill, which is important
for oil combating, will suffer as a consequence.
Oil spill modelling in ice covered water is not widely described in the literature
(Reed and Aamo 1994) although there exists a certain pool of experiments (Yapa
and Belaskas 1993) and observations (Wang et al. 2008). The wind drift is reduced
in the presence of sea ice depending on the ice concentration (Leppäranta 2010).
The process is discussed in Johansen et al. (2005), Drozdowski et al. (2011). Sea ice
influences oil drift in dense ice (ice concentration larger than 70 %) while there is no
influence in low ice concentrations (less than 30 %). For oil at the surface, the drift
velocity is a combination of current speed and ice drift velocity, with proper weights
assigned depending on the ice concentration. This is a rather crude approach, limiting the effect of the ice to the drift velocity. weathering processes are significantly
slower in ice leads, but the spatial scale is of the order of meters, which is typically