SPECIAL SERIES | DEEP IMPACT
Deciphering the unseen, underwater effects of the Gulf oil spill.
What viewers might have missed, however, was that the images were brought to them by a technology that went remarkably right—deep-sea robotics.
That same tech allowed humans to drill in such deep and dangerous waters in the first place, then helped plug the leaking well, and is now contributing to groundbreaking research on the impacts of the BP spill.
During the spill—as workers skimmed slicks, lay booms, and cleaned oiled beaches above—muscular minisubs known as work-class ROVs (remotely operated vehicles) battled a mile (1.6 kilometers) below the waves to control the worst underwater oil spill in U.S. history.
At the wellhead linked to the BP-leased Deepwater Horizon rig, which sank on April 22, ship-based operators had the robots trying to activate the busted blow-out preventer, cutting pipe, unbolting equipment, attaching heavy hoses and sensors, and, of course, taking that high-def video.
“Then the phase went into the choke and kill operations, where we had to intervene, cut, modify, adapt, all underwater, all using the ROVs,” said James Miletech, ROV supervisor for Oceaneering International, in a video on the company’s website.
Oceaneering, the largest ROV operator in the global oil patch, was helping Transocean and BP drill the well before the Deepwater Horizon exploded, and afterwards was instrumental in capping the runaway well. “What’s normally done on the rig floor with impact wrenches, people on ladders, on work platforms, using powerful tools, we had to modify these tools to work subsea,” said Miletech.
Oceaneering engineers had to design and build hundreds of custom tools, virtually overnight, one of which cut away the bent riser pipe. Others helped install the hoses that allowed thousands of barrels of oil and gas to be siphoned off by a drill ship. Still other robots reactivated a sensor on the damaged blowout preventer that told how much oil was flowing and how much pressure was really there—a critical number to BP engineers and federal officials.
Whirring Into the Plume
While ROV crews worked virtually nonstop to control the Gulf spill, an entirely different set of underwater robots was deployed to monitor its unseen spawn: a deep, morphing plume of toxic hydrocarbons suspected to be creeping through the Gulf, and perhaps beyond.
Among the monitoring 'bots was one of the Spray AUVs (autonomous underwater vehicles) developed by the Scripps Institution of Oceanography in California.
Known as a glider, because it has no propeller, the 6.5-foot-long (200-centimeter-long), torpedo-shaped Spray dives or rises by shifting its buoyancy, thus gliding on a roller coaster-like trajectory through the water column.
With their low energy requirements, the battery-operated Spray gliders can cruise for months at a time. On board, each Spray carries basic oceanographic sensors that measure conductivity, depth, and salinity. Each glider also hosts an acoustic Doppler profiler, to measure currents, and a tool that uses light to measure fluorescence as a way to detect phytoplankton—or oil.
The Gulf Spray AUV spent most of the summer away from the spill site. The glider's main job was to monitor the eddy that splinters off the Loop Current (diagram) and into the central Gulf.
Rising every six hours or so to transmit its data via satellite phone, the AUV allowed the U.S. Naval Oceanographic Office to produce nearly real-time maps of oil in Gulf currents.
"The great advantage of gliders is that they have the ability to measure and predict ocean conditions right away," said Dan Rudnick of Scripps, who controlled the glider via his laptop and iPhone from his office in San Diego. "You want to know where that spill is going right away."
Such real-time knowledge of currents can help responders defend against spills—for example, by quickly dispatching boats to burn, skim, or absorb the oil or by moving endangered turtle eggs out of harm's way.
Robotic Bird Dog of the Deep
Another AUV was sent specifically to hunt for the deep-sea oil plume: a Dorado-class self-propelled AUV, developed by the Monterey Bay Aquarium Research Institute (MBARI) in California.
Also torpedo-like, a Dorado is larger—up to 21 feet (6.5 meters) long—moves faster, and has more instruments than a typical glider. These MBARI AUVs were originally intended to look for sediments coming off the continental shelf or from river deltas—making Dorados ideal for detecting organic matter, such as oil in the deep Gulf of Mexico (map).
About three miles (five kilometers) southwest of the broken BP wellhead, the Dorado spent two days in June creating a high-resolution visual and chemical snapshot of a section of the Gulf measuring 2.2 miles long by 0.9 miles wide (3.5 by 1.5 kilometers) and between depths of 3,000 and 4,000 feet (900 and 1,200 meters).
The AUV's sensors detected consistently high levels of hydrocarbons in the plume. The Dorado also took water samples, which were analyzed for hydrocarbons back in the lab. So far the samples seem to corroborate what the sensors found.
The oil-content data is critical, because many of the heavier, more toxic compounds are also among the most persistent.
The Dorado used in the Gulf illustrated research robots' increasing ability to "think" on their own.
For the plume mission, MBARI scientists developed software algorithms that allow the Dorado to change its behavior. Rather than simply "mowing the lawn"—back and forth, back and forth—the AUV, by gauging fluorescence, could pinpoint areas rich in colored dissolved organic matter (CDOM), including broken-down crude, and whirr in to collect samples.
"When I give presentations on this, I use a picture of a bird dog," said MBARI oceanographer John Ryan, whose team deployed the Dorado in the Gulf. "The continuing developments in software will only make these robots smarter."
Taking the Plume's Measure
One of the "smartest" robots operating in the Gulf this summer was Sentry, built by Woods Hole Oceanographic Institution in Massachusetts.
Shaped like a sunfish, or mola, the skinny, slab-sided, 5-foot-tall (150-centimeter-tall) submersible swims upright, carrying an array of oceanographic equipment. (See pictures of other animal-like marine robots.)
Sentry's acoustic sub-bottom profiler, for example, can examine sediments 15 feet (5 meters) beneath the seafloor. And a mass spectrometer allows the robot to take real-time measurements of hydrocarbons in the water column.
Deployed in late June about three miles (five kilometers) south of the spewing BP wellhead, Sentry's critical task was to determine the subsurface plume's shape, direction, composition, and provenance—part of the Natural Resource Damage Assessment required under the U.S Oil Pollution Act, which the federal government is to use to determine how to penalize BP for the spill.
Crisscrossing the plume countless times and recording tens of thousands of data points, Sentry documented a plume 22 miles (35 kilometers) long, 1.3 miles (2 kilometers) wide, and 650 feet (200 meters) tall at a depth of about 3,600 feet (1,100 meters).
"It sounds big," said Richard Camilli, lead scientist on the Sentry team. "But in the vastness of the Gulf of Mexico, it was looking for a needle in a haystack."
The plume "gets exponentially harder to find the farther you get from the source," Camilli added. "And it's a moving target, meandering around, not following a continuous path in time, like a stream of smoke in the air."
One of the important discoveries made by Camilli's team was that oxygen levels in the plume were only slightly lower than normal—leading researchers to conclude that microbes weren't quickly breaking down the plume, as some had suggested. (Read more about how nature fights the oil spill.)
Camilli hopes to take Sentry back down to the Gulf to look for the heavier types of oil and gas that tend to sink to the bottom and can persist for years.
Next: Robotic Rigs?
It's not just spill-response missions that are benefitting from rapid advances in robotics. The technology stands to speed undersea oil prospecting and perhaps prevent some drilling disasters from occurring in the first place.
For example, the Norway-based Seabed Rig AS company is working on a drill rig similar to the Deepwater Horizon. But unlike that BP-leased rig, Seabed's design would essentially be a giant seafloor robot.
Part of what made Deepwater Horizon's job so risky were the little-known threats of drilling at such tremendous depths, under massive water pressure. But the incentive to go deeper isn't going away, said Neil Tardella, whose Massachusetts-based company Energid Technologies is helping to develop software for the Seabed rig.
"An estimated 40 percent of untapped oil in the world is in deep water or in the Arctic—very difficult places to get to," Tardella said.
And while robotic rigs may not make drilling at depth exponentially easier or less expensive, they could well make it safer, Tardella added. "If an exploratory rig like this [had been] built," he said, "no one would have died in the BP disaster this summer."