Researchers at the Massachusetts Institute of Technology and Shell Oil's international exploration and production arm said they have been examining such channels, which are filled with highly permeable and porous sedimentary deposits that extend deep below the sea floor.

The structures form when sediment-laden currents flow off the outer continental shelf and into channels on the deep-ocean floor, depositing layers of sand, silt and clay as they go. Over millions of years, the channels can become filled with porous sandstone covered by impermeable mud � creating a perfect trap for oil and gas that seep up from below.

Typically, companies recover only 30 percent to 40 percent of the oil in a given reservoir. Because a single reservoir may contain 1 billion barrels, increasing recovery efficiency by even a single percentage point would mean a lot of additional oil.

Over the past 20 years, energy companies have withdrawn significant amounts of oil from such buried channels, but they could extract even more if they understood the channels' internal structure.

"If we could understand how they develop, then we would also understand a great deal about what they're composed of � the distribution of clay, silt, sand and even gravel that they're built out of," said co-researcher David Mohrig of MIT.

With a better understanding of porosity and permeability within a channel, Mohrig said, companies could determine more accurately how much oil is present, where it is located and how quickly it can be withdrawn.

Mohrig's team has been re-creating the formation of the submarine channels in his Morphodynamics Laboratory using a 5 meter-square sand table.

The experiments have yielded results the researchers call counterintuitive. On a map, the sinuous submarine channels look like meandering surface rivers, but they exhibit behaviors that are markedly different and unexpected.

The behaviors stem from differences in density. Water in a river is about a thousand times denser than the fluid it flows through - air. As a result, a flow tends to remain confined to its riverbed, escaping over the banks only rarely.

In contrast, the current running through a submarine channel may be only 10 percent denser than the seawater around it, so the current can spill out of its channel more easily and frequently than a river might.

That difference explains several findings, Mohrig said. For example, at times the bottom of the current sloshes almost all the way up the edge of the channel and then back down again. At bends, the current may go straight, pouring up and over the bank and dropping its sediment outside the channel - an outcome with important implications for energy companies as they plan to drill.

"The experiments that David is doing have never really been done before, so we're learning new things about how channels are put together," said co-researcher Carlos Pirmez of Shell Oil. "We're getting new ideas, new concepts that may change the way we think about the subsurface."

The result, he added, should be improved predictions, reduced uncertainty and more efficient recovery from these oil-rich submarine formations. Related Links
MIT Geomorphology Lab