After a major oil spill, government agencies and energy companies have limited time to make decisions on how to remediate the disaster and clean up the spilled oil. Officials currently rely on computational models that predict the behavior of the oil plume in order to determine the best strategies for containing and cleaning up the oil.
While these models provide insights on the spill that are crucial for rapid response cleanup efforts, existing simulations don’t account for some processes that play important roles in determining the direction, shape and overall size of the oil plume.
A professor at the UH Cullen College of Engineering is improving the computational model used to predict the behavior of oil plumes in deepwater blowouts, such as the BP Deepwater Horizon disaster in 2010. Once deployed, the improved models will provide insights to more effectively remediate these kinds of oil spills.
Di Yang, assistant professor of mechanical engineering at the Cullen College, and his collaborators at Johns Hopkins University and UCLA received $877,162 from the Gulf of Mexico Research Initiative (GoMRI) in support of this research. A total of $247,490 will support Yang’s portion of the research at UH.
In order to glean as much useful information as possible about an oil spill within an hour or less, officials rely on rapid response models, or integral plume models, that simulate only the average behaviors of an oil plume in the near-field of a leaking oil well to predict the amount of oil spilled, how long it will take for the oil to reach the ocean’s surface and in what direction the oil plume will likely travel.
But there are many small-scale interactions and processes that can have a profound impact on the overall behavior of an oil plume. For instance, the sizes of the oil droplets and gas bubbles released from the site of an underwater blowout have an impact on the shape of the near-field plume and its interactions with the ambient sea water. Small-scale turbulence at the edge of the oil or gas plume also plays a vital role in the mass and momentum exchange between the plume and the ambient sea.
Due to time and computational power constraints, this information is either averaged together or filtered out of simulations in integral plume models.
“The rapid response model has critical value for quickly determining how to respond to an oil spill, but the accuracy of this prediction heavily relies on the incorporation of other physical models to supplement the rapid response model. This accounts for the effects of some critical physical processes that are filtered out due to averaging,” Yang said.
Yang and his colleagues are using Large Eddy Simulation (LES), a high-fidelity computational simulation technique for modeling turbulence in fluid dynamics, to fill in the gaps of the current integral plume models. The physical models developed by Yang’s team can be seamlessly integrated into existing integral plume models without requiring additional computational power or time to receive the results.
“Using the computer power and resources currently available, we will directly resolve as much information as possible that the current integral models can’t resolve,” Yang said.
Yang began this research as a postdoctoral researcher at Johns Hopkins University in 2013. With funding from GoMRI, Yang’s team studied the complex physical systems at play when spilled oil travels across the surface of the ocean.
Results of the research showed that not only large-scale systems – such as climate and ocean flow – play important roles in the dispersion of oil, but also small-scale interactions between oil droplets and upper-ocean turbulence. His team also found that the use of dispersants – chemicals used to break oil into smaller droplets so that it biodegrades more readily in the ocean – can profoundly impact how and where an oil plume travels across the ocean’s surface.
Once the underlying physics of these interactions were quantified, Yang’s team used their findings to supplement the large-scale ocean circulation models used in predicting regional dispersion of an oil spill in Gulf of Mexico.
“There’s a big gap between the smallest scale resolved by ocean circulation models and the scales at which the oil plume and ocean turbulence interact, so that’s what we’re filling in,” he said.
Next, Yang and his colleagues focused on modeling the behavior of an oil plume at the site of a deepwater blowout. In a study published in the Journal of Fluid Mechanics last May, Yang’s team simulated a lab-scale multiphase plume that mimicked the dynamics of the Deepwater Horizon plume as it burst vertically from the well at the ocean floor. The team then used a numerical model they developed to understand the underlying physics of the oil plume’s behavior.
In the case of the BP Deepwater Horizon Blowout in 2010, government officials approved the company to apply dispersant directly at the site of the burst well, nearly 1,500 meters below the surface of the ocean.
“It reduced the damage a lot, but it’s been questioned how efficiently that strategy actually worked,” Yang said. “We want to provide insights to find better remediation strategies for responding to these kinds of oil spills in the future.”
Using LES, Yang’s team advanced the understanding of the fundamental physics and physical processes that contribute to the behavior of the oil as it shoots upward from the well at the ocean floor.
“The more we understand the fundamental physics, the more accurate our integral models will be. Now, when the rapid response model spits out averages, there will be fundamental physics underlying its results,” said Yang.
With the current funding from GoMRI, Yang’s group will take their research a step further by integrating both the vertical and horizontal model of oil dispersion into a complete framework.
“Right now, if a company or the government wants to look for a model that can capture all of the essential flow physics to provide a precise prediction of how an oil plume will behave during its entire fate – from the bottom of the ocean to the surface – we don’t have that kind of model,” Yang said. “After we understand the true physics and integrate these models together, we will be able to predict with greater accuracy how an oil plume will behave, and that will lead to better remediation strategies for future oil spills.”