Following the torpedoes, the next major advance in the technology of wrestling oil from rocks occurred in 1932, when Dow Chemical began to use hydrochloric acid to dissolve rocks and create channels for oil. The first test took place in Midland, Michigan, near Dow’s headquarters. Company engineers mixed 500 gallons of acid with arsenic, used to prevent the steel pipes from corroding. It worked, increasing the well’s flow by threefold. The next year, in North Texas, another company decided to try injecting acid more forcefully into a well. This time 750 gallons of hydrochloric acid were injected, followed by oil to force the acid into the limestone formation. Before this treatment, the well had yielded only 1.5 barrels of oil a day. After the acid treatment, it flowed 125 barrels a day. A new industry was born. By 1938, some 25,000 wells had been “acidified,” and individual wells sometimes were given as much as 10,000 gallons of acid. But acidization had limited usefulness. It worked in limestone but not sandstone. This limitation was problematic, since many oil and gas reservoirs are sandstone. Engineers were on the lookout for something else to open up the rocks and force wells to give up their riches.
One of the centers of the industry’s research effort was Tulsa, the self-proclaimed “Oil Capital of the World.” The heavyweight in town was Stanolind, which possessed girth and swagger from its origin as one of the companies formed in 1911 when the Supreme Court broke up the John D. Rockefeller’s Standard Oil trust. Soon after World War II, Stanolind undertook one of the earliest efforts to gather engineers and scientists on a campus to improve the industry’s knowledge and develop new techniques to maximize production. In 1952 it opened a palatial new research center on sixty acres at the cost of $4.5 million. The local newspaper marveled that it creates “a country club atmosphere, so immaculate does it keep its lawns.” Only a couple of companies had independent research units, including the Texas Oil Company, later Texaco; and Standard Oil Company of New Jersey, later Exxon. The industry’s early decades were full of wildcatters who relied on superstitions and luck to find oil, but after World War II, the largest oil companies wanted to rely on science and engineering instead. (Some superstitions were unknowingly rooted in geological insights. Some early prospectors drilled in church cemeteries, believing that oil accumulated underneath. Their instinct was right, although not their logic. Many cemeteries were built on hills, and these high spots were an expression of a rock fold or salt dome underneath. Later, geologists realized that these subterranean structures often contained oil and gas.)
Riley “Floyd” Farris, a studious man committed to applying science and mathematics to improve oil-field operations, emerged early on as one of Stanolind’s star researchers. He graduated from the University of Oklahoma in 1935, nearly a decade before the school awarded its first bachelor degree in petroleum engineering. His interest was in cement. Then, as now, drillers deploy cement to fill the empty space in a well, locking the pipe in place. If the well is drilled through a shallow zone of water, the cement helps keep the water out of the well and anything from leaking into the aquifer. Farris wrote several papers trying to exhort manufacturers to provide better-quality cement and operators to pay more attention to how high temperatures and pressures degraded cement.
One oil-field mystery intrigued Farris. Drillers noticed that when cement entered a well, some of it occasionally disappeared. Cement wasn’t cheap, and when it was lost, the well required additional costly cementing. Some wells needed up to five batches of cement before they sealed properly. (At least, that was what good practice dictated. Sometimes it was deemed good enough and work proceeded. Some states had regulations covering how to cement wells; others didn’t.) The lost cement puzzled Farris. Why did some wells take more cement than the amount predicted by his slide-rule calculations? What was happening? Where was the cement going?
While this phenomenon had been noted before, Farris was the first to study it systematically. He pulled files from 115 wells and determined there was a mathematical relationship between the pressure created by the cement and the depth of these wells. His conclusion was straightforward. The weight of the cement and other liquids in the well were rupturing the rocks, creating fractures. When the cement was squeezed into the wells, some was being lost into the cracks. What if he
tried
to fracture the rock by pumping in liquid? Unlike cement, liquid could be removed after it cracked the rocks. Once the fluid was removed, he thought, perhaps more oil and gas would seep out of the rocks.
While Farris was quiet, his colleague Bob Fast was his opposite. He was friendly and outgoing, at ease both in the laboratory and in the oil field. His father worked as a tool pusher, or manager, of drilling rigs in Illinois. When oil discoveries in the Midwest petered out, the family migrated to Tulsa. In the booming city, oil-field work was easy to find. Fast received a degree in petroleum engineering at the University of Tulsa in 1943 and then spent a year fixing cracks in the wings of Douglas SBD Dauntless dive bombers during World War II. A year later, he joined Tulsa-based Stanolind Oil’s brand-new research effort.
Fast was witty and had a devilish sense of humor. After many years with Stanolind (which became Amoco and later part of BP), he purchased land on the shore of the nearby Lake o’ the Cherokees to indulge his love of boating. He bought a double-sailed sloop and named it the
Dammit Virginia
after his second wife. He designed the lake house himself, burying it half underground to conserve energy and, in the 1970s, covered it in solar panels to provide heat and hot water. He also installed solar panels atop his home in Tulsa. His son, Rob Fast, remembers neighbors gawking at their unusual roof, then alien to Oklahoma. But Fast loved his solar panels. “It was free energy. It was practical technology and economic,” said his son.
In November 1946 Bob Fast set out to test Farris’s theory in the Hugoton natural gas field in southwestern Kansas. Fast was a twenty-five-year-old looking to make a name for himself in Stanolind’s research department, and he hoped that fracking the Klepper #1 well would provide a wanted career boost. Colonel Roberts performed the first frack jobs using explosives, but Fast and Farris pulled off their first fracks using a liquid. Since water generates friction—requiring a lot of pumps to inject it into the well—Fast looked for a way to reduce water’s friction. He needed a liquid that was slick, mixed well with water, and was readily abundant. Fast settled on napalm left over from World War II, since it was no longer needed to fuel flamethrowers and fill bombs dropped over Japan.
Worried about fire hazards, the mixing tanks and pumps were placed about 150 feet from one another, creating an odd sight in an industry where equipment was typically crowded together. The unusual well configuration caught the attention of industry spies. Numerous companies employed scouts to keep an eye on their competitors to see how deep they were drilling and if the wells were successful. In the pancake-flat area of Kansas, the spies didn’t even try to hide. They often parked near a well and watched the operation, taking notes dutifully. The Chevron scout assigned to the Klepper noted the size of the pipes used and the well’s depth. “Rust” was all he wrote, cryptically, under drilling remarks.
Fast pumped in one thousand gallons of napalm-thickened gasoline, followed by two thousand gallons of gasoline. He repeated this four times at different depths. He appears to have created fractures in the limestone. The pressure dropped each time, indicating that the liquid was leaving the well. When the napalm and gasoline were recovered, gas flowed out of the well. But it was about the same amount of gas that would be expected with a conventional well that had been soaked in acid. The frack job was a failure, but Fast and Farris would soon return with other experiments.
Stanolind’s research into what it called “hydrafrac treatment” wasn’t in the name of pure science. The company wanted to make wells more productive. By the middle of the twentieth century, the difficulty of finding large new oil fields weighed on the industry. It settled for lesser wells and focused on cutting costs and wringing every barrel out of a well. Oil companies had already invested in drilling these wells and building pipelines, so a new technology to get another 5 percent of the oil or gas in the reservoir could be quite profitable. What’s more, during World War II, steel was needed to build tanks, bombers, and other military machines. A relatively shallow 4,300-foot-deep well required sixty-one tons of steel for the pipes. The oil industry was caught in a bind. It needed to increase oil production to satisfy the war effort, but it had limited supplies of steel required for new wells and pipelines. The solution was to return to oil wells that had already been drilled but were languishing. By the 1940s, the industry had drilled more than one million wells. More than half had been abandoned altogether or were producing just a trickle of oil and gas. These wells were considered played out. Fast and his colleagues at Stanolind wanted to see if fracturing wells could resuscitate old wells and make new wells more productive.
The United States ramped up oil and gas production to meet demands of fighting World War II. After the war effort wound down, demand for fuel kept rising as the economy boomed. By 1948, there were 3.3 million more cars on the road than seven years earlier. There were nearly a million more oil furnaces heating homes. With demand outpacing production, the industry was running full tilt to keep up. In February 1948 cold weather spread across the country, and fuel shortages occurred. A Chrysler plant in Detroit laid off workers because it couldn’t get enough natural gas from Texas. There wasn’t enough fuel oil for furnaces, and people lined up in Chicago and Saint Petersburg, Florida, to fill jerry cans. Even before this episode, the industry realized the growing urgency to deliver more oil and gas. Its search took it offshore, and in 1947 the Oklahoma City energy exploration company Kerr-McGee made the first oil discovery from a platform in the Gulf of Mexico—so far offshore that it was out of sight of land. One of Kerr-McGee’s partners on the well was Stanolind Oil.
While Kerr-McGee was making history offshore, Bob Fast and Floyd Farris kept pursuing their idea that onshore wells could be fracked. The problem they faced was that while they could infer what was occurring when cement was lost, they didn’t really know. They wanted to see what happened. So they drilled a nine-and-a-half-foot well into a shallow sandstone formation near their Tulsa offices and squeezed in cement. Then they dug up the well to see what had happened. What they saw confirmed Farris’s hunch. The cement was fracturing the sandstone and spreading out in all directions. In one area, it had traveled more than five feet from the well. It may have spread even farther, but they had excavated only a five-foot circle. The result was so striking that they repeated the experiment a dozen more times. Fast and Farris found cement that had caused vertical as well as horizontal fractures.
As they grew comfortable with their ability to fracture rock with cement, they encountered a problem. Fractures with water would close up once the water was removed. Why not mix in some sand with the water to prop open the fractures? This was first tried in East Texas on a well that was producing less than a barrel of oil every day. A mixture of sand, crude oil, and a soap laced with metals was pumped into the well and left to sit for forty-eight hours. The soap scrubbed the oil off the rocks. When the petroleum concoction was removed, the well began to produce fifty barrels of oil a day. The engineers hoped their hydrafrac treatments would boost a well for a few weeks or even a few months. But to their surprise, the fracked wells kept flowing. It wasn’t the oil industry’s answer to eternal youth, but it was a revitalizing drug that kept aging wells producing like teenagers for a few years. They had widened the well’s drawing radius, sucking more hydrocarbons out of the ground.
In May 1948 Farris filed a fracking patent. “This invention pertains to a method of increasing the productivity of an oil or gas well by providing lateral drainage channels in selected formations adjacent to a well,” he wrote. The patent includes a description of an East Texas well. Before it was fracked, it was barely flowing. The well produced only enough oil to fill a tablespoon every eight seconds. Fast, as usual, conducted the field experiment. About 122 barrels of fluid were injected into the well: a mixture of crude oil, solvents, and aluminum soap—the latter an ingredient in napalm, now used for waterproofing. He mixed sand into the liquid, hoping that it would remain in the fractures, propping them open. The liquid was pumped in to 3,400 pounds per square inch of pressure, about the same as a top-of-the-line, commercial-grade pressure washer available today. The sandstone fractured. Liquid flowed in. Fast kept the liquid in the well for two days and then took it out. On a sustained basis, the well flowed fifty times more oil than it had before being fractured. An exclusive license was issued to HOWCO, the Halliburton Oil Well Cementing Company.
For decades, the industry had poked holes in the earth, praying to get lucky and hit a gusher. It had even found some success with brute force—setting off nitroglycerin in a well—to get out more oil. The Stanolind experiments hinted at a different future. Engineers could manipulate the earth. Deeply buried rocks could be conquered by the engineers being churned out of universities. It was a turning point for the oil industry, even if it wasn’t obvious at the time. The age of the wildcatter was drawing to a close. The age of the petroleum engineer had begun. From this point on, the industry would be defined by men convinced they had the tools and science to bend rocks to their will.
In October 1948 one of Fast’s colleagues wrote up Stanolind’s findings and published them in the profession’s leading journal,
Transactions of the American Institute of Mining Engineers
. The paper sets out the basic elements of modern fracking: pumping in liquids under pressure to create a fracture and sending in sand to prop open the cracks. The hydrafrac process boosted production in eleven out of twenty-three wells. And it wasn’t expensive. “It is significant that the value of the additional oil and gas produced to date through the benefits of this process has already exceeded the combined cost of research, development and all field tests,” the paper noted. The work at Stanolind was an immediate sensation. Journal editors made it their lead paper for the year.