143
three- dimensional computational fluid dynamics (CFD) models, based on large
eddy simulation (LES, Fraga et  al. 2016). Since CFD modeling approaches are
computationally costly to implement, comprehensive models designed to guide
response during an oil spill still use the integral modeling approach (Johansen et al.
2003; Zheng et al. 2003; Gros et al. 2017; Dissanayake et al. 2018).
The initial conditions to each of near-field models we consider include the oil
and gas composition, physicochemical properties, flow rate, orifice characteristics, water column conditions, and depth. Thermodynamic calculations partition
the released petroleum fluid into gas and liquid phases and predict their properties
throughout the near-field region. These models consider the main fate processes
affecting the gas bubbles and oil droplets. Dissolution is the primary fate process
in the near-field, and gas dissolution should be modeled to accurately predict the
intrusion location (Socolofsky et al. 2015). Dissolution from liquid-phase petroleum can also be significant for the fate of light hydrocarbons, especially for deepwater releases of live oil (Lehr and Socolofsky 2020; Oldenburg et  al. 2020;
Malone et al. 2020, Pesch et al. 2020; Gros et al. 2016, 2017; Dissanayake et al.
2018). Other important processes to consider are heat transfer and the density
equation of state of gas and liquid petroleum as the oil composition changes and
the pressure and temperature evolve. For breakup, equations are needed to predict
interfacial tension and viscosity. Each of these processes is significant during the
rapid cooling near the release and in the early stages of dissolution. Once the oil
and gas droplets reach the intrusion layer, however, much of the rapid dissolution
has completed, and biodegradation begins to compete with dissolution for the
dominant fate processes.
When the petroleum fluids exit the orifice and enter the ocean, break-up models
simulate the formation of bubbles and droplets. To predict the evolving sizes and
complete size distribution throughout the turbulent jet, physics-based models are
needed (Zhao et al. 2014, 2015, 2017). These physical models compare the resistive
forces from interfacial tension and viscosity for each bubble and droplet in a flow to
the destructive forces from turbulence that work to break the gas and liquid into
bubbles and droplets. These models respond to the rapidly varying turbulent dissipation rate in a plume and have been developed to handle gas, oil, and the effects of
injected subsea dispersant, both in the near-field, turbulence-dominated regime and
in the tip-streaming-dominated breakup, occurring in the buoyant plume, and potentially in the intrusion layer and far-field domain.
Following the momentum-dominated jet break-up region near the release, a
plume develops that is controlled by the collective positive buoyancy of the oil and
gas, the negative buoyancy of the entrained, stratified ocean water, and the drag
resulting from the lateral currents. Integral models simulate these effects by tracking the trajectory of a control volume, integrated laterally across the plume and
tracked along the plume centerline. Eulerian (Jirka 2004) and Lagrangian (Lee and
Cheung 1990; Lee and Chu 2003) integral models exist, with all blowout models
being of the Lagrangian integral plume model type. In these models, entrainment
occurs by shear entrainment at the plume edge (Turner 1968) and by engulfment of
9 Dynamic Coupling of Near-Field and Far-Field Models
Précédent

- 154/617

Suivant