Three Iowa State University engineers—one drawing on hands-on experience and the others applying theory, mathematics, and computational modeling—offer a detailed look into how the Deepwater Horizon spill in the Gulf of Mexico may have begun, and how planners can prepare to minimize the effects of any future spills.
Cement expert questions judgment, not technology
Bob Steffes knows about deepwater oil rigs and he knows about cement, and even though he doesn’t yet know all the details about what happened in the Gulf of Mexico, this much seems likely: in an effort to save time and money, poor decision making, not poor technology, doomed the Deepwater Horizon.
“We know how to do this,” he says of drilling for oil thousands of feet below the ocean surface, and many thousands of feet further into the earth. “It’s safe as long as everything is done properly, and most of these materials have been around for decades. This isn’t something new and wild out there that we can’t handle. Corners were cut and it bit them.”
Steffes, a PCC (Portland Cement Concrete) research engineer at Iowa State University’s Institute for Transportation, bases his conclusions on 17 years of overseas oil rig experience, including a blowout on an offshore well in the Middle East.
His explanation of how an offshore drilling operation works sticks to the basics, and it’s the attention to those basics that begins to make some sense of the situation as the public is flooded with media reports and frustrated by the dribble of information coming from BP.
While no simple explanation may ultimately emerge to account for the exploratory well’s failure, a number of factors, in the opinion of Steffes, played a role. Many of them relate to “loss circulation” – a common problem in this sort of drilling, in which drilling material intended to be circulated back up the drill pipe instead filters out into the surrounding porous rock. In this case, the loss involved the drilling mud (water and clay with dense additives such as barite or hematite) that we’ve all heard so much about.
“When I heard that loss circulation was a problem, I knew right away” what might have contributed to the blowout, Steffes said. He watched the incident carefully from the first day and has been compiling information along with the rest of us, although with a more expert eye.
What does he see?
- The chemical wash used to clear away the “filter cake”—that is, the drilling mud that accumulates on the face of the drill hole through loss circulation—was not completely effective.
- Only a few (perhaps only six) devices known as casing centralizers, which keep the final casing in line within a series of casings, were used when more than 20 should have been used. That may have resulted in the final cement seal being off-center, which can allow gases to escape.
- Nitrified (or foamed) cement was used instead of a heavier, stronger version to seal the bottom of the well. Steffes points out that the cement itself, although of low strength, wasn’t the issue, but that the amount that was used may have been insufficient because it, too, may have been lost in the same way as the drilling mud.
- No bond log was performed. This process uses sensors to determine if there is a uniform sheath of cement between the outside of the casing and the drilled hole. Gaps, if detected, are routinely filled – a “squeeze” operation, according to Steffes.
- Drilling mud was replaced by seawater. Typically, the entire pipe would be filled with mud, some of which would then need to be cleared away once the exploratory rig was replaced by a production rig. In this case, the top portion was filled with seawater. The net effect of this time-saving step reduced approximately 180,000 pounds of overhead load.
- The full-length casing that was used provided only a bottom hole cement seal. Using a liner would have provided a bottom hole cement seal plus a liner hanger seal.
Steffes notes that none of this would have mattered if the blowout preventer had not failed, and learning more about why this one did will someday shed even more light on the disaster.
Researchers bring element of control to spill’s complexities
Throughout the still-unfolding oil spill episode in the Gulf, one theme has predominated: uncertainty. How much oil is flowing? When will we stop it? Where will it go? How can the spill be contained?
By studying what is happening now, two Iowa State researchers hope to use mathematical equations and computational models to provide tools ahead of the next potential disaster—tools that account for uncertainty but emphasize control.
Umesh Vaidya, assistant professor of electrical and computer engineering, is an expert in dynamical systems and control theory, and has extensive experience in dealing with complex systems and their control. Baskar Ganapathysubramanian, assistant professor of mechanical engineering, uses mathematics and high-performance computing to model real-world phenomena, where uncertainty is a given.
The two young professors had already found that their knowledge blended well in research collaborations (for example, control of complex fluid flows). But as the oil spill emanating from the Deepwater Horizon’s exploratory well began to expand, they saw opportunity in the midst of tragedy. An opportunity to offer the world insights for keeping the next spill under control.
“It’s very important to understand the natural flow of a system,” Vaidya said. “My work involves developing theoretical models and tools that will allow me to understand the behavior of a complex system and ultimately control its behavior.”
In this case, the complex system is the ocean currents in the Gulf. The two researchers have gathered data from the National Oceanic and Atmospheric Administration so they can build a model of the currents.
“Once you have that model,” Vaidya said, “you analyze and try to understand it, and then try to design a control mechanism to allow you to change the behavior of the system. The goal,” he said, “is to understand how ocean currents behave naturally and then take advantage of the natural dynamics for the purpose of control.”
By having an intricate understanding of the currents, Vaidya explained, it’s much easier to know where chemical dispersants should be sprayed for maximum effect, or to determine the optimal location for booms and other barriers. This can be achieved, he said, even though conditions can change by the day or even by the hour.
“Right now we’re at the stage where we’re actually developing the models and doing simulations to verify if the output represented in the model is realistic,” Vaidya said.
That’s where Ganapathysubramanian’s expertise comes in. Through the use of partial differential equations, he must find ways to account for uncertainties in weather patterns, wind flow, and the uncertainty of the oil flow itself. By using the parallel computing resources at Iowa State, he can simulate a system with multiple uncertainties. Doing so allows him to consider potentially all possible scenarios.
While early results may be available within a month, the longer-term goal is to create software packages that could be placed online, thus making them available as resources that could be used immediately when the next spill occurs.
“The message that we want to get out is that we really need to understand the natural dynamics and the uncertainties in the ocean flow currents first, before there is a spill,” Vaidya said. “Our work points to where the oil might go and when it might get there. We may even find that there are natural barriers to the oil flow that we can take advantage of and learn where those barriers are.”
Eric Dieterle, College of Engineering, (515) 294-4881, email@example.com
Bob Steffes, Institute for Transportation, (515) 294-7323, firstname.lastname@example.org
Umesh Vaidya, Electrical and Computer Engineering, (515) 294-7975, email@example.com
Baskar Ganapathysubramanian, Mechanical Engineering (515) 294-7442, firstname.lastname@example.org