The Boom (21 page)

Before Devon even closed the deal for Mitchell, it sent some of its engineers to embed with the shale operations and get a better idea of what was going on. Brad Foster, head of midcontinent operations for Devon, was given control of the Barnett Shale. He grew up in Pittsburgh and, in the late 1970s, during summers while in college in West Virginia, he worked for a local gas company that tried a couple fracks into shales. He has two recollections from those summers. The first: “How does anyone make any money doing this?” The second: “I will never work a shale project again.” Just the guy to be handed the Barnett Shale.

Foster faced a tricky situation. Mitchell Energy had been drilling a lot of plain vertical wells, straight down, and then using its new fracking technique. This practice worked well enough in a portion of the Barnett Shale where the rock sits above limestone. But there was a much larger area where there was gassy Barnett, but no limestone. Underneath the shale was the Ellenberger formation, rocks riddled with salty water. When wells in this area were fracked, the slick water injected into the well came in contact with the salty reservoir. Most of the energy of the frack—the massive horsepower assembled to force in millions of gallons of water—would dissipate in the Ellenberger. Instead of creating large man-made fractures to drain the Barnett, it would create small fractures into the Ellenberger. The result was expensive salt-water wells. If Devon wanted to create a replica of the Dead Sea in northern Texas, these wells were ideal. For any other purpose, they were duds.

Devon had a different approach in mind, one that Mitchell had gingerly attempted years before with federal research grants. The idea was to drill horizontal wells in the Barnett. A vertical well is like an elevator that picks up passengers at every floor. The Barnett is typically 350 feet thick, or the size of a thirty-five-story building. Devon’s horizontal wells would traverse through the shale, running a couple thousand feet. This elevator, laid on its side, would be longer than the world’s largest skyscraper. Devon’s horizontal wells would reach much more of the Barnett gas from a single well. There was another potential major advantage to horizontal wells. Devon hoped that these wells would help it avoid the Ellenberger’s salty trap. If the horizontals didn’t work, Nichols knew he had paid a hefty price for Mitchell. But if Devon could frack where the Barnett sat above the Ellenberger, the number of wells it could drill on Mitchell’s acreage would triple or quadruple. Buying Mitchell would turn out to be a steal.

One day in 2002, Nichols and Foster sat down to talk about plans for the Barnett. “I want you to expeditiously but carefully see if these horizontals work,” Nichols told Foster. He warned, “Don’t bet the family farm.” Once a month, Foster’s new Barnett team gathered in an unadorned conference room on the seventh floor of a downtown boxy building with a gray, bland lobby that looked more like it housed a collection of small accounting firms than the global headquarters of a company listed on the New York Stock Exchange. The only indication that the building housed an oil company was a model in the lobby of an offshore drilling platform.

Members of the Barnett team brought maps and printed-out spreadsheets that tracked costs and gas production trends, as well as the time to drill and complete wells. Foster has thick fingers and a balding pate that make him look like Tony Soprano, but he has a warm personality. He cracked jokes and emphasized the need to experiment, but to try new approaches carefully. It “was about as far from full steam as you could get,” said one attendee. Soon Foster turned the talk to horizontal wells as a way to get more out of the Barnett. “Why can’t we bring this technology here?” he asked. Many of the field engineers, the people who actually drilled the wells in the Barnett, were holdovers from Mitchell Energy and resisted the idea. They had tried these wells and didn’t think they would work. Some worried that an expensive failure would blot their careers. “Here’s the deal,” Foster told the team. Referring to Devon’s management, he went on, “There are no repercussions. We are going to take all the accountability on this, and we’re going to take the responsibility.”

So, Foster continued, what are the obstacles? An engineer responded that there were no rigs in the Barnett that could handle these kinds of wells. A few days later, Foster and a couple members of his team drove to Tulsa to meet with Helmerich & Payne, a rig outfit. He told the company that he wanted a rig capable of horizontal drilling in the Barnett. “They looked at us with their eyebrows raised,” he said. “What are you up to?” an H&P executive asked. Foster explained his theory. If the well could be turned to run directly through the Barnett Shale, perhaps the fractures wouldn’t escape into the Ellenberger and produce salt-water wells. “We think it will work,” Foster said. H&P agreed to send one of its most modern rigs to Fort Worth on a trial basis.

Horizontal wells, in 2002, were fairly unusual. Only one of every fourteen wells drilled in the United States and Canada was horizontal. (A decade later, six of every ten wells were horizontal.) While these twisting wells were still relatively new, the industry had been drilling slant wells for decades. In 1941 oil was discovered under the Oklahoma Capitol building. Even in oil-crazed Oklahoma, tearing down the steel-reinforced concrete dome was out of the question. Instead, Phillips Petroleum set up a drilling rig across the street. The well proceeded at an angle to travel underneath the building, and for years lawmakers met atop an active oil well. The well, nicknamed the Petunia #1, was drilled from a flower bed.

In 1986 Al Yost, the government scientist in West Virginia, and colleagues drilled a horizontal well into the Devonian Shale in southwestern West Virginia’s Cabwaylingo State Forest. This well descended for 2,113 feet vertically, and then, using pipes bent at a slight 2.5 degree angle, the drill bit proceeded slowly to the right. The pipe got stuck once, and the angle achieved was too small, forcing them to retreat a few feet and start a new shaft. The motor that drove the drill plugged up several times and broke down every ten hours, on average. But over the distance of nearly six football fields, they reoriented the well until it ran horizontally, parallel to the ground for 2,000 feet. Viewed at a remove, the well resembled a truncated capital J. The well dipped into and ran through the shale, exposing more of the rock to a rudimentary frack. Yost reported in a paper that after fracking it, the well flowed at a rate seven times higher than vertical wells in the area.

Since that West Virginia well, the industry had gotten better at horizontal drilling. It can be done faster and with more precision. Devon drilled its first horizontal Barnett well in June and July 2002. The Veale Ranch #1H took nearly a month and a half to drill and was fracked with 1.2 million gallons of water. The well worked, although production was only marginally better than a vertical well. Five months later, in November 2002, Devon had its sea legs under it and was gaining both confidence and speed. Devon drilled its sixth horizontal well, the Graham Shoop #6, in less than a month. It used more than twice as much water as the Veale well to frack it, and more than seven times as much gas came out. In Foster’s eyes, it was a beautiful well. His misgivings about drilling shale began to melt away. But when he asked his team at the monthly Barnett meetings to explain to him why the wells were so good, blank stares met him.

Shale wells are overachievers. The new fractures free up a lot of natural gas, which will rush into the well and up to the surface. The first few weeks that the wells are connected to a pipeline yield the highest production, and then they start to decline. A well will continue producing gas long after the gas freed up by the initial fractures travels up into the well. This gas appears to come from the shale rock itself, worming its way out and into the fractures. How this works remains not entirely understood.

Hundreds of engineers filled out an informal survey at a shale gas conference in November 2008. “I am confident that I understand reservoir drainage” was one question. Four-fifths of the engineers responded that they weren’t confident. Several years into the all-out juggernaut of shale drilling, the experts themselves didn’t know exactly why shale production worked. They just knew it worked.

If these engineers were puzzled about what was going on in the shales, what hope did Brad Foster have? In early 2003, he started to get production reports from Devon’s first horizontal Barnett Shale wells. They looked promising. But he wasn’t ready to pop any champagne. “We were a little bit on the pessimistic side,” he said. “We were not sure we totally understand this shit.” The world’s first modern frack well, the S. H. Griffin #4, was all of five years old at that point. It had started off producing more than one million cubic feet a day but had dropped off to a quarter of that by mid-2003. Who knew where it was going?

Larry Nichols said later that the company was acting with what he characterized as “excessive conservatism.” The lack of data led Devon to inch forward, instead of break into a run. “What we wanted to see was, Okay, it peaks, and we know it goes down at a very steep rate; where does it flatten out? If it flattens out at one level, you make a lot of money on that tail. Or does it continue on down and flatten out at a lower level?” he explained. Until time passed and he had real production data, he decided to “tread water” for a couple years.

Devon was following a time-tested method: drill a little, wait a little, drill a little, wait a little. In technical terms, the company was asking whether the decline curve would be exponential or hyperbolic. Exponential decline means that production falls off until it just peters out. Drawn on a graph, it looks like a straight line. Hyperbolic decline features a dramatic drop followed by a flattening out. If the declines were hyperbolic, as Foster thought they would be, then shale gas could be profitable under the right circumstances. But until he knew for sure, neither he nor Larry Nichols wanted to risk too much money drilling the Barnett.

In June 2002 Devon held a Barnett coming-out party for Wall Street analysts in a Dallas suburb. There were 1,043 wells in the Barnett, it explained, but there was room for another five thousand. It was using fourteen rigs in the Barnett. At that pace, it had a fifteen-year backlog of wells to be drilled. Not only wasn’t Devon in any hurry, it had planned exploration around the world that would take a lot of money to finance. Ramping up spending in the Barnett wasn’t a top priority. There was no need, Devon thought, to go out and find other shales similar to the Barnett. There weren’t any. The company declared that it was “unique.” It couldn’t have been more wrong.

That fall, I decided to take a drive around the Barnett to see for myself what I had been hearing about in corporate presentations. Back then, you had to go looking for signs of gas development. The rigs were few and far between. I drove to a slight rise so that I could gain a vantage point to see above the strip malls and chain restaurants and find a rig. I saw one off in the distance and spent the next half hour taking wrong turns down cul-de-sacs until I found it. It had been erected in a field a couple hundred feet behind a cluster of suburban houses. Neighbors said it was loud. That it was lit up like an oversized Christmas tree at night annoyed them. This type of close proximity between homes and drilling pads was becoming the new normal. For the first time since the early 1900s, when oil extraction in Los Angeles began, an urban drilling campaign was under way.

In July 2003 Devon offered me a tour to show off the Barnett. I met a couple company employees at a Shell station off Interstate 35. Nearby, several trucks idled, waiting their turn to fill up from a metered municipal fire hydrant. We visited a drill pad and a large plant that took the gas and stripped out ethane, butane, and propane. Fracking hadn’t entered the national discussion. The tour was a way for Devon to talk about its horizontal wells and demonstrate how it was getting gas out of the ground. Fracking wasn’t a curse word. Not yet, anyway. Devon didn’t even call it shale gas back then. My tour guide, a former Mitchell employee named Jay Ewing, called it “unconventional gas” to separate it from the normal way of drilling for the fuel. Years later, I met him again and spent another few hours driving around the Barnett. On the second tour, he showed off the three-story-tall sound walls that Devon built around its drilling pads to cut down on the noise and be a better neighbor. This time he told me his full name. My first official tour guide to the new world of fracking was none other than J. R. Ewing.