Graduation Year

2023

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Mechanical Engineering

Major Professor

Andres Tejada-Martinez, Ph.D.

Committee Member

Rasim Guldiken, Ph.D.

Committee Member

Wenbin Mao, Ph.D.

Committee Member

David Murphy, Ph.D.

Committee Member

Ping Wang, Ph.D.

Keywords

Langmuir Supercells, Oceanic Turbulence, Oil Spill, Population Balance Model, Resuspended Sediments

Abstract

Langmuir turbulence in the inner shelf coastal ocean is characterized by Langmuir circulation (LC) or cells that can span the full depth of unstratified water columns. A Reynolds-averaged Navier-Stokes (RANS) simulation strategy resolving full-depth LC coupled with an oil-particle aggregate (OPA) formation model is introduced. The RANS methodology is employed in two different cases. The first case corresponds to a shallow shelf zone of uniform depth, focusing on vertical mixing and transport of OPAs. The second case pertains to a surf-shelf transition zone with variable depth, comprising vertical and horizontal mixing of OPAs. In the latter case, a wave model that accounts for the wave effects on current (WEC) is coupled with the Reynolds averaged approach to determine dominant wave characteristics such as amplitude, wavenumber, and direction as functions of cross-shore location. The phase-averaged surface gravity wave model is equipped with dissipation due to wave-breaking and bottom drag to account for wave effects on currents.

It is seen that full-depth LC generated by wind and waves under storm conditions can result in sediment resuspension and oil droplet entrainment (in the case of an oil spill) and subsequent mixing between these two, leading to significant OPA formation. This conclusion comes from the shallow shelf zone of uniform depth simulation in which various classes of oil droplets (with diameters ranging between 40 and 140 μm) were released at the surface; each class was initialized with a 0.1 kg m-3 concentration. The full-depth LC led to most of the oil becoming trapped within OPAs in the first 5 minutes. The majority of the larger oil droplets were quickly aggregated with sediments near the surface, whereas the smaller oil droplets were first submerged by the downwelling limbs of the Langmuir cells and eventually aggregated with sediment while being carried upwards by the upwelling limbs. The OPAs conglomerated in the form of clouds transported by the action of the cells, with the heavier OPAs eventually settling within the upwelling limb of the cells while slowly depositing to the bed over time.

In the modeled surf-shelf transition zone, a wind-driven circulation with the effect of Langmuir turbulence is obtained where the wind direction is set perpendicular to the shore (favoring onshore transport). The wind stress generates a mean cross-shore flow with surface current directed onshore and a compensating undercurrent directed offshore. Furthermore, the wind and the waves generate LC oriented in the downwind direction perpendicular to the shore. As the waves approach the shore, their amplitude rapidly decreases, particularly within the surf zone (nearshore), due to wave breaking, resulting in a reduction of Langmuir cells intensity in this region.

The turbulence (via the turbulence model), the surface and sub-surface currents, and the resolved Langmuir cells impact the transport and formation of OPAs. Oil droplets were released at a distance ~800 meters from the shore on the surface of the water. The initial concentration of the spilled oil followed a Gaussian distribution spreading over 50 m (covering a few LCs). The peak oil concentration of this distribution was 0.1 kg m-3. Simultaneously, a static homogenous spatial sediment concentration is prescribed with a mass concentration of 0.02 kg m-3 per sediment class. The surface current transported oil droplets and OPAs toward the shore, while the subsurface current transported submerged scalars offshore. By ~30 minutes, all the oil had stopped being consumed towards the formation of OPAs. Larger oil droplets aggregated at a faster rate (~8 mins.) than the smaller ones.

Initially, OPAs form a cloud near the surface that follows the shape and size of the initial oil spill. Over time, the cloud is broken apart and carried by multiple LC downwells. Due to their settling velocity, the larger OPA classes reach the near-bottom region of the water column quicker, where they may be advected offshore by the mean subsurface current. The transport towards the shore was greater for the smaller OPA classes since these remained within the surface current for greater times. Furthermore, the uniform spatial distribution of sediments used in the simulations favored the generation of medium sized OPAs (180-360 μm). This is the first simulation that has explicitly computed (resolved) LC in nearshore coastal regions and its impact on the transport of oil and the formation of OPAs.

Download

Share

COinS