Energy companies are only "scratching the surface" of U.S. gas shales production, which is expected to grow 160% by 2020, said Scott Reeves, executive vice president of oil and gas consulting firm Advanced Resources International.

Gas shales wells typically have long reserve lives and low production rates and capture gas over a wide geographical area rather than in discrete structural traps. The gas molecules are "adsorbed," or held to the surface of the organically rich marine shales, which were typically produced in the Devonian period.

Reeves cited the gas shales success of Mitchell Energy & Development Corp. in the Fort Worth Basin of north central Texas as an important element in enticing Devon Energy Corp. (DVN) to acquire Mitchell in January.

He stressed that the economics of Mitchell's gas shales play in north central Texas has improved considerably over time.

Before 1990, Mitchell had drilled 75 wells in the area at a cost of $600,000 to $1 million a well. They had an estimated ultimate recovery per well of 0.3 billion cubic feet to 0.5 billion cubic feet, and finding and development costs of $1.20 to $3.33 per thousand cubic feet, Reeves said.

By 2000, though, Mitchell had used new technology to better define the reservoir, space wells more closely, and produce more per well. By 2000, Mitchell was spacing one well per 27 to 55 acres, versus one well for every 320 acres before 1990.

The closer spacing translated to economies of scale, profitable production, and much higher booked reserves. By 2000, the cost per well had dropped to $400,000 to $600,000, estimated ultimate recovery had risen to 0.8 billion cubic feet to 1.2 billion cubic feet of natural gas per well, and finding and development costs had sunk to 33 cents to 75 cents per thousand cubic feet.

Reeves also cited Burlington Resources Inc. (BR) for its gas shales efforts in the San Juan Basin of New Mexico.

The use of better technology is key to making gas shales production a profitable proposition, according to Reeves. Better technology is enabling "sweet spot" exploration in areas of natural-fracture clusters, where the permeability of the rock is improved because of natural fractures, often near fault lines.

Reservoir characterization and modeling, using three-dimensional seismic information, is also key, as is proper stimulation of the rock to enhance production.

Of the total U.S. gas production of 19.2 trillion cubic feet in 2000, only 8.8 trillion came from conventional onshore production. Another 4.8 trillion cubic feet came from conventional federal offshore production, and 5.6 trillion cubic feet came from nonconventional production.

Conventional production is expected to decline between now and 2020, while nonconventional production is expected to grow by 59%, Reeves said. "We think there's a very, very bright future for nonconventional gas."

Of the nonconventional production in 2000, 3.7 trillion cubic feet came from tight gas sands, 1.4 trillion cubic feet came from coalbed methane, and 0.5 trillion cubic feet came from gas shales. But gas shales production is expected to grow the fastest of the three.

"Tight gas sands" are found in the lower to middle parts of basins and are extensive and don't have much water. "Coalbed methane" production occurs from coalbeds with low-cost, shallow wells after water over the coalbeds is pumped out, enabling the gas to escape from the coals.

By 2020, gas shales production is expected to grow to 1.3 trillion cubic feet per year, or 160% higher than the 0.5 trillion cubic feet recorded in 2000. Tight gas sands production is expected to grow to 5.7 trillion cubic feet per year, 54% higher than 2000 production, he said. Coalbed methane production is expected to grow to 2.1 trillion cubic feet by 2010, and then drop down to 1.9 trillion cubic feet per year by 2020, still 36% higher than 2000 production.

"Gas shales is a real opportunity," said Reeves. "We think that the Rockies will be the focus of much of that growth."