Simulation of a Subsurface Oil Spill

OIL SLICKS IN THE OCEAN: PREDICTING THEIR RELEASE POINTS USING THE NATURAL LABORATORY OF THE SANTA BARBARA CHANNEL

Principal Investigators: Jordan Clark (UCSB), Bruce Luyendyk (UCSB), and

Summary of Research

The main purpose of the project is to understand the role played by seep bubbles in the transport of hydrocarbons including oil from the seabed to the sea surface. The goal is to validate a numerical bubble model to better predict the surfacing footprint of oil, thereby improving spill mediation efforts and preparedness.

Numerical studies showed sensitivity to several parameters, including size, seep depth, upwelling flows, and saturation of the plume water. Since these parameters are largely unknown in the literature, our approach has been to measure these parameters at a very active seep site, Shane Seep, as well as at several other seeps in the seep field. In the process several discoveries were made.

Bubble Measurements

Measurements of the size distribution at the seabed are used to initialize bubble models and thus are highly critical. Bubble distributions were measured at the seabed at Shane Seep and published in Leifer and Boles (2004). They showed that seeps can be loosely classified into four categories, major, minor, obstructed, and elastic. Elastic vents were observed at Ira Seep, a tar mound that transiently released very large pulses of bubbles, with bubbles up to 20 cm diameter. Obstructed seeps are where a vent is obstructed physically, such as by a log or anemones, and tended to have very large bubbles. Minor vents produce bubble streams at a low enough flow rate that the size distribution is very narrow, and were in the size range of 2000-3500 µm radius. Major vents produced bubble flows sufficiently strong that bubble breakup occurs, and have a broad and weakly size dependent bubble size distribution.

Fluid Motions

Dye release experiments had quantified upwelling velocities in the plume and provided dramatic evidence of the existence of the upwelling flow. From the bubble distributions, upwelling velocities were inferred for different bubble streams and are published in Leifer and Boles (2004). Direct measurements of fluid motions by dye releases are published in Clark et al. (2003) and were ~30 cm/s for Shane Seep.

Seabed Morphology

During the last few years, numerous changes were observed in the seabed morphology and in areas of active seepage (i.e., vent activation and deactivation, crater growth and shrinking). These changes were identified as related to transient seepage events that were observed both directly and through air pollution records. A theory on the effect of tar on gas seepage was developed from these observations. Changes in the seabed features and the theory were published in Leifer et al. (2004a).

Oil Emissions

Bubble distributions and upwelling flows were analyzed and interpreted and showed a variability that could not be explained absent bubble oiliness. It was discovered that at major vents, very oil bubbles would occasionally be produced by the breakup of large bubbles. These bubble-oil droplets rose very slowly, following a different trajectory than the vast majority of bubbles. From minor vents, occasional very oily bubbles escaped from the vent mouth. It was believed that a 4 hz oscillation in bubble emission resulted from the interaction between oil and gas flow through the vents, resulting in a cyclical variation in oil/gas ratio. These results were published in Leifer and Boles (2004).

Geochemistry

Gas samples collected and analyzed have suggested that Shane Seep emissions have an unusually high CO2 concentration (~12%). Measurements were then made of the CO2 profile in the first 3 m above the seabed, and showed a rapid exponential decrease. Surface bubble gas composition was measured for deeper (70 m) and shallower (20 m) seeps and showed more air content at the deeper seeps. Water concentrations were also measured in the bubble plumes at the deeper and shallower seeps. Deeper seeps showed super saturations not measured at the shallower seeps. This was interpreted as the the bubble stream in water column had not had sufficient time to reach equilibrium with only 20 m water, but did reach saturation in 70 m water. These results were published in Clark et al. (2003).