IN DECEMBER 2003, an international team of geologists announced that they had successfully tapped a new energy source. Methane hydrate, a solidified form of natural gas bound into ice, lurks under the seafloor along the margins of every continent and under the Arctic permafrost. On the Mackenzie River delta in the Canadian Northwest Territories, engineers drilled hundreds of meters below the permafrost into the hydrate deposits. They punched fractures into the layers of sediment and pumped hot water into the earth, releasing the natural gas from its icy prison.

This first harvest of methane hydrate could mark a new direction for the energy industry. Engineers once assumed that the energy costs of melting the frozen fuel would outweigh the gains. But rising oil and gas prices and creative uses of existing technology, like the recent test in the Canadian Arctic, are beginning to change their minds. The United States Geological Survey estimates that the total amount of natural gas in methane hydrates surpasses all of the known oil, coal, and gas deposits on Earth in energy content, although only a fraction of the frozen fuel will be extractable. The hydrates can form at any latitude on Earth if temperature and pressure conditions are right, and are usually mixed with sediment under the ocean floor.

There is a catch, however. Methane hydrates offer the energy industry dangers as well as opportunities, warns Charlie Paull, a geochemist at Monterey Bay Aquarium Research Institute in Moss Landing, Calif. Deep-sea drilling operations that melt seafloor deposits of the icy fuel might set off an underwater accident under certain circumstances.

The hazard results not just from tapping into hydrates themselves, but from oil companies’ and governments’ drive to explore for petroleum in deeper waters than ever before, Paull says. Propelled by the highest oil prices in a more than a decade, engineers in the Gulf of Mexico and the North Sea are extracting oil and natural gas in waters more than a kilometer deep — entering the zone where methane hydrate mingles with sediment and rock.

Normally, the pressure of hundreds of meters of water above keeps the frozen methane stable. But heat flowing from oil drilling and pipelines has the potential to slowly destabilize it, with possibly disastrous results: Melting hydrate might trigger underwater landslides as it decomposes. Scientists hypothesize, in fact, that 8,000 years ago, decomposing hydrate helped to generate a gigantic landslide under the North Sea. The resulting tsunami scoured the Norwegian fjords and scattered seafloor sediment across Holland and Scotland. While no one is predicting that drilling could catalyze an event of such catastrophic proportions, an underwater slide in an oil field could cause enormous environmental damage from oil spills that couldn’t be easily stopped.

More controversially, another danger of the frozen hydrates arises from the fact that methane is a potent greenhouse gas. Some geologists have suggested that methane could accelerate global warming if the oceans’ rising temperature eventually released the gas in large-enough amounts. Rapid, methane-driven global warming has occurred before in Earth’s history, causing mass extinctions, they say, and humans could make it happen again if we keep warming up the planet with the exhaust from cars and electricity plants.

In the near future, what experts such as Paull worry about most is the risk from oil drilling. As the energy industry proceeds into deeper waters in search of fresh oil and gas deposits, Paull says, it has neglected the hazard that melting methane hydrates might pose to its own infrastructure A single $1 billion offshore platform can house 100 people and withstand hurricane-force waves and winds, but Paull suspects that with a big enough nudge from below, pipelines could break.

“Those oil platforms are some of the largest and most expensive structures ever constructed by humans,” he says. “The chance of an incident is very small, but can we afford to have just one? The oil industry has not addressed scientists’ questions about seafloor stability to my satisfaction in a public way.”

Paull is closely familiar with the double-edged nature of methane hydrates because he has been studying them for 15 years. In 1996, when he worked at the University of North Carolina, he ran the first drilling trip dedicated to looking at methane hydrates. He and his colleagues demonstrated the presence of methane hydrates on a part of the Atlantic sea floor off the Carolina coast called the Blake Ridge. Hunting for the icy deposits, in the last five years, he and MBARI scientists have taken samples of sea floor sediment from locations around the world.

Recently, he went to the Gulf of Mexico to map the hydrates and assess their risk to the oil industry. And this summer he will travel to the North Sea to investigate the seafloor at the site of the ancient landslide, where energy companies are developing a huge oil field. Each destination tells a different story about the frozen fuel that can help researchers assess whether methane hydrate is an energy boon, or a disaster waiting to happen.

IN ALL THE PLACES Paull has investigated, the same process that generates methane from swamps, and from human intestines after a meal of baked beans, also supplies the main source of the gas for hydrate formation. Bacteria in ocean mud close to continental coastlines feast on organic material in the sediment and belch out their exhaust. Caps of relatively impermeable hydrates sometimes sit above and trap reservoirs of free natural gas.

To make hydrates, water molecules link up in a cage-like structure — resembling ice — with small “guest” molecules such as methane sticking between them. One cubic meter of methane hydrate is packed with the equivalent of over 160 cubic meters of methane. It melts at room temperature and atmospheric pressure. When the water above is deep enough, methane hydrate’s zone of stability extends from the sea floor down to where the internal heat of the Earth starts to warm things up.

Just like natural gas, methane hydrate would burn more cleanly than coal or oil. The DOE forecasted in 2003 that the world’s natural gas consumption will grow the fastest of all energy sources in the next 25 years. Governments expect that with increasing demand, research into techniques for recovering methane hydrates will pay off in a couple decades.

Until recently, the energy industry mainly regarded methane hydrates as a nuisance. Engineers on oil platforms regularly confront the frozen deposits because they form spontaneously from cold water and gases flowing through pipelines, sometimes plugging them for weeks or even causing blowouts. The $12 million that the U.S. Department of Energy has allotted to research on harvesting hydrates since 2001 is small compared with the estimated $100 million U.S. firms spend every year on antifreeze, repairs, and other gas-flow—assuring remedies.

“It’s an interesting flip,” says Richard Charter, a marine conservation specialist at Environmental Defense in Oakland, Calif. “In the past, the oil industry did everything in their power to avoid disturbing hydrate deposits,” says Charter, who calls himself the “token environmentalist” on a federal advisory board on hydrate research. “Now, it’s a potential resource. Oil engineers’ eyes get really big when you start talking about it.”

In 1999, a USGS report estimated that the world’s free natural gas deposits could yield 368 trillion cubic meters of the methane. By comparison, the report approximated that U.S. offshore areas contain over 10,000 trillion cubic meters of gas in hydrate form. Geologists have since adjusted both figures downward, says USGS scientist Timothy Collett, but the more recent figures still give no sense of how much gas could actually be produced from hydrates. The Mallik test itself produced 1500 cubic meters a day, enough energy to serve about a thousand American households, although a small amount compared to nearby natural gas production.

The most promising places to mine hydrates, he says, are sites where deposits are concentrated, like veins of ore — such as in the Arctic. But the Gulf of Mexico is also a hot target. The Gulf already accounts for 30 percent of U.S oil production and the bulk of exploration for new oil reserves. “The crucial thing about the Gulf of Mexico,” Collett says, “is that when we figure out how much methane hydrate there actually is, the infrastructure to take advantage of it already exists.”

CHARLIE PAULL AND A GROUP of MBARI and USGS scientists spent two weeks in 2002 in the Gulf of Mexico on the French research ship Marion Dufresne. They were there to map methane hydrate — and assess the potential for a landslide triggered by oil drilling. “Ten years ago, we asked: Where can we find methane hydrate?” Paull says. “Now, it’s more: How can we figure out where it is not?”

In the Gulf, geologists have found rich hydrate deposits bursting through the seafloor sediment in mounds a few meters wide. Tubeworms and mussels feed on the methane. Some of it comes from bacteria in the mud, but the gas also is constantly seeping up from pressure-cooked organic material deep within the Earth. To protect these rare delicate ecosystems, federal government regulations prohibit drilling near the seafloor mounds.

In the Gulf, drillers and operators have previously avoided areas where hydrates are close to the surface. “There is a geohazard. It’s worth considering and preparing for,” says Tom Williams, an engineer at Noble Corporation in Houston, which operates mobile offshore drilling units for the oil giants. To lessen the risk, companies can use double casings with refrigeration while drilling to make sure the sediment around a pipe doesn’t heat up.

Williams points out that drillers have bored through hydrate-rich sediment many times in the Arctic with little incident. However, he also says there have been oil-well blowouts in Alaska that some geologists have blamed on hydrate.

Charlie Paull says the danger of geological instability is probably less in the Arctic, compared with other places, because there, methane hydrate is encased by hundreds of meters of sediment or sand. In the Gulf, by contrast, hydrates can lie close to the ocean floor.

Geologists and oil engineers agree that more information about where hydrates are located is essential, but taking measurements of the stuff in seafloor sediment is a challenge, as Paull and colleagues found on their 2002 Gulf trip. Oceanographers initially estimate the locations of hydrate deposits by probing with sonar, or the scientists find them by direct observation. The underwater pressure that stabilizes methane hydrates can get in the way of detailed study. Working from the Marion Dufresne, the MBARI-USGS team took piston cores, giant cylinders of seafloor sediment up to 50 meters long and 10 cm wide, from 21 locations around the Gulf. To release pressure from the free gas produced by decomposing hydrate, they poked holes in the cores along the plastic casings. “Mud worms” of grey goo came squirting out. The pressure can blast sediment out of the top of the core barrel, flying in one instance 10 meters into the air before landing in the water. As soon as the concentration of methane gets high enough to be interesting, Paull says, “it fizzes out like Alka-Seltzer — making your measurements meaningless.”

One possible solution is to keep the core pressurized as it is brought up to the surface. But cumbersome machinery limits the number of samples taken this way, and the cores extend only about a meter long. Instead of measuring methane in cores directly, Paull and Ussler have developed a different method to gauge the amount of gas diffusing up from lower deposits. They look for falling concentrations of chemicals such as sulfate, which is consumed by methane-eating bacteria in the sediment. In seafloor mud, the concentration of sulfate and other chemical signatures are proxies for methane below: The less sulfate, the more methane.

The MBARI-USGS team will soon publish the results of their research: At the bottom of the Gulf of Mexico, methane hydrate is only present in small amounts away from the ocean ridges. The risk of structural instability on the seafloor mounds still exists, Paull says, but the team’s work can begin to assure everyone that there isn’t a large-scale danger to the coast extending from Florida to the Yucatan.

Meanwhile, the search for deeper deposits in the Gulf of Mexico is slowly revving up. A DOE-funded project to drill for hydrates in the Gulf, headed by oil giant ChevronTexaco, is scheduled to begin in 6 to 12 months. One of the proposed sites for drilling is directly under a region of the seafloor that geologists know to be rich in hydrates.

THE OTHER HOT SPOT for investigating methane hydrate is in the North Sea. This summer, Paull and colleagues from MBARI will travel there to probe the potential of the frozen fuel deposits to generate a monster landslide. They will take cores from the site of the ancient slide that occurred 8,000 years ago, when a cliff off the coast of Norway at a place called Storegga collapsed. The initial slide generated a wall of water 15 meters high and moved an amount of rock comparable to submerging the state of New Jersey.

Geologists disagree about the prominence of methane hydrate’s role in the Storegga slide. According to one theory, the slow warming of the ocean after the end of the last ice age melted hydrates embedded in sediment under the steep part of the cliff. This made the sediment like a plate of Jello, ready to slip away when the right earthquake came along, according to a recent review by U.S Navy marine scientists. But Norwegian scientists say that pressure from layers of sediment settling from above drove the slide, and that the temperature below the North Sea at the time of the slide is uncertain.

At Storegga, Paull and colleagues want to reconstruct the past. Their research, in cooperation with scientists from the University of Wyoming and the University of Tromso in Norway, will try to answer the questions: How much gas escaped from the ancient landslide, and how much methane hydrate still lies below? Methane leaves a particular chemical signature of carbon and sulfur in the mud that the MBARI scientists plan to analyze.

Their investigation touches present-day developments. The Norwegian state energy company Norsk Hydro is planning production from a $9 billion gas field called Ormen Lange, which lies in the middle of where the Storegga slide occurred. Norsk Hydro will begin building the world’s longest subsurface pipeline this year to deliver gas from Ormen Lange; it will climb up the sloping sea bed to Norway and then across the North Sea to the United Kingdom.

Norwegian research predicts there will be only one major slide at a given site for each ice age, says Martin Hovland, a geologist at the Norwegian state oil company Statoil, a partner in developing the Ormen Lange field. Norsk Hydro and Statoil’s internal data suggest that because of the escape of hydrate deposits after the Storegga slide, the sea floor needs a full climate cycle over tens of thousands of years to recharge sediments with methane. “We take the geohazard issue very seriously of course,” Hovland says. “We’ve performed very stringent drilling and sampling over the last six years, and have come to the conclusion we can develop the field without major problems.”

Paull is not convinced. “While I suspect they have a strong case, it is frustrating because much of the data on which that conclusion has been reached is not yet publicly available,” he says, communicating by e-mail.

A Storegga-scale slide could have consequences beyond making big waves. Some geologists think the amount of methane released in the Storegga disaster could have been enough to warm up the earth. Methane is 20 times more powerful a greenhouse gas than carbon dioxide. In a 1991 Geophysical Research Letters paper, Paull, Ussler, and USGS geologist Bill Dillon proposed that when glaciers suck up too much water during an ice age, the sea level drops enough to release the pressure on methane hydrates all over the globe. Enough methane could escape into the atmosphere, they hypothesize, to trap the sun’s heat and end the ice age.

The idea that such a release of methane hydrates played a role in mass extinction events in Earth’s history is gaining acceptance. Geochemists at University of California, Santa Cruz, recently reported evidence in Science that methane fuelled a 5 to 10 degree rise in the Earth’s temperature 55 million years ago. About a third of all species of a common marine plankton perished, and the heat drove an exodus of early mammalian species across the continents. Evidence also exists for a similar event 600 million years ago.

James Kennett, a geologist at University of California, Santa Barbara, who discovered evidence of the 55-million-year-old extinction event, warns that the melting and release of methane hydrate on a global scale “is the right mechanism to propel climate change,” he says. “It happens very fast geologically — over a few decades. But the climate fluctuations in the last few thousand years are small compared to the big events 55 million years ago.”

Paull’s collection of data on the fate of methane released in the North Sea landslide at Storegga could provide evidence to help evaluate Kennett’s hypothesis. But scientists such as Roger Sassen, a geochemist at Texas A+M University, are skeptical of the theory. Based on his own observations — taken from submarines — of constant leaks of methane into the Gulf of Mexico, Sassen views methane hydrates as a trap for huge volumes of greenhouse gas and a buffer against climate change. He also says that their harvest poses no climate risk because freshly generated methane is always seeping up from below, and using it will just capture what already leaks into the atmosphere.

Geologists may argue about the feasibility of harvesting methane hydrate and its influence on climate, but Paull and Sassen agree on one thing: that deep-water drilling will require a different kind of thinking from the oil industry. Adequately dealing with the hazards of working in deep water and eventually harvesting and transporting methane hydrate, they both suggest, will require new technology that does not yet exist.

“When drilling, oil engineers usually just roar right through the zone where hydrate is stable,” Paull says. “We really have to develop a new set of tools to even know what’s going on in extremely deep water.”