The spectacular blowout in 1901 at ‘Spindletop’ in southeastern Texas ushered in the modern oil industry and greatly changed the world’s energy mix. People realised the potential of oil to surpass coal as an energy source, leading to the discovery of huge deposits of conventional oil and gas around the world. While the Barnett Shale discovery started in a much quieter way, its impact is no less important and has transformed oil and gas exploration and exploitation.
Against the Odds
While the Spindletop and Barnett discoveries look very different on the outside, they share some interesting similarities, both beating the odds against them. These included:
Each very small for their time without the large financial and research resources of a major.
Determined oil men:
A self-taught geologist named Patillo Higgins tried to drill the salt dome at Spindletop for more than ten years. The first wells failed and Pennsylvania oil men Galey and Guffey were persuaded to finance another well. Similarly, George Mitchell, a Texas A&M educated geologist and petroleum engineer, spent millions drilling hundreds of wells over 17 years before finding economic success.
Sceptics and the unknown:
Before the Spindletop discovery, most people in the petroleum geology business thought Higgins’s ideas of finding oil in a salt dome were nonsense. Similarly, little was known about the potential of producing gas from shales. Mitchell’s own board of directors were sceptical and considered his efforts in the Barnett a waste of money.
While drilling the Spindletop discovery well, drilling difficulties and cave-ins forced the crew to pump mud instead of water down the hole to remove cuttings, a revolutionary idea at the time. The mud worked by coating the drill hole, allowing quick progress when the well had to be deepened. The Barnett discovery got its breakthrough when engineers switched from gel to water fracks. The change in fluids used made a major difference for both discoveries and those innovations are still used today.
Changing an industry’s outlook and future:
Spindletop blew oil nearly 50m into the sky and caused a stampede of drilling and new discoveries all over Texas and the world. Once Mitchell proved the economic success of the Barnett, the oil and gas industry’s outlook on unconventional reservoirs greatly changed and prompted the drilling of thousands of new wells in shale basins across the US and other countries.
In 1973 the Arab oil embargo prompted a great awakening across the US; our oil and gas supplies were dwindling and we could be vulnerable to supply disruptions. As a result, the Department of Energy (DOE) was formed in 1977, unifying energy planning within the federal government by putting a host of federal energy-related agencies and programmes into a single, presidential cabinet level department.
The government started to look to new sources of energy and unconventional gas was one of them. The US gas industry began back in 1821 near Fredonia, New York, where a gas well produced a few thousand cubic feet each day from Devonian shales continuously for 35 years. This seemed a natural place for the DOE to start to support research and development of unconventional resources. The Eastern Gas Shales Project (EGSP) was initiated in 1976, a year before the DOE existed, but soon became an integral part of the department’s Unconventional Gas Recovery Program. The project promoted commercial gas development from Devonian shales in the Appalachian Basin, which could contain up to 900 Tcfg. They also studied advanced drilling technology such as directional drilling and advanced stimulation technology.
North Texas: the First Sixteen Years
While this was happening Mitchell Energy was busy drilling wells in North Texas. Mitchell’s investment in the area was substantial and it needed a new gas source to reverse the company’s declining gas production, so George Mitchell began thinking about the Barnett Shale. About this time Dan Steward joined Mitchell as part of a new team of geoscientists and engineers. Dan became the coordinator between the Barnett team members as “a lot of people in the company thought the Barnett was a waste of time and money and did not want to get their hands dirty on the project”.
“In 1981, George had Slay No. 1 drilled to evaluate shallow conglomerates and the deeper Viola Limestone which lies below the Barnett,” recalls Dan. “The Viola turned out tight and Mr. Mitchell had always wanted to test the Barnett. While drilling many previous wells, the Barnett section would give up good gas shows. We were not very optimistic at the time but went ahead with fracking and testing the Barnett. After stimulating with N₂ (no sand or water, based on the EGSP’s work with Devonian shales), the well produced 120 Mcfpd. It was refracked the next year using CO₂ foam, which increased gas production to 274 Mcfpd.” (Note: The Slay well has been fracked five times and is still producing.)
The next 40+ wells Mitchell fracked and tested in the Barnett would also test the company’s fortitude. Dan Steward explained, “We started out with small fracks and poor results, finally going to much larger gel fracks consisting of 400,000 gallons of water and 1,250,000 pounds of sand. By 1991, Mitchell had a good position in the basin but we were still lacking a knowledge base. That was when George Mitchell said he was open to bringing in DOE and the Gas Research Institute (GRI). We started by evaluating Barnett core and then GRI sponsored a horizontal well in the Barnett.” GRI’s assessment of that horizontal well concluded, “The Barnett Shale drilling and completion economics favour hydraulic fracturing in vertical wells.” Mitchell drilled two additional horizontal wells, but the bulk of their producers were vertical.
“From 1987 until 1997, 304 Barnett wells were all stimulated with these massive hydraulic gel fracks,” Dan continues. “Most of these wells were deepened down to the Barnett from wells testing shallower objectives which really improved the economics and were also helped by gas contracts and federal tax credits for tight gas. We had wells that produced up to 1 Bcf ultimate recovery. It was near the end of this testing period when we realised we did not have open fractures in the rock.”
“Geophysics had a pretty small part in the early development of the Barnett,” says the project’s geophysicist, Jon Huggins. “However, our mapping showed that as we got near faults, production would drop off. Initially we thought open fractures were needed to get results, but it turns out that was not necessary. We finally shot a small 3D survey and found the area much more complex than mapped from the older 2D surveys. It turns out, with the more recent switch to horizontal drilling in the play, detailed seismic mapping is necessary to accurately locate wells.”
Success at Last
The massive gel-fracks were very expensive, most costing more than the actual drilling, but they got the job done. However, there were other problems besides the cost. Most of the reservoirs being fractured were either tight or at such low permeabilities that the formation was not able to ‘clean-up’, and the fractures had much shorter effective half lengths than designed.
At this time, engineer Nick Steinsberger was running the fracking effort in the Barnett and experimenting with different liquids and gels to create pathways for the gas to escape. “I had been charged with cutting fracking costs to improve the economics of each well,” says Nick. “With wells only producing 1 Bcf and the gel-fracks costing in excess of $375,000, something needed to be done. For a couple of years we kept experimenting with reducing the gel loads. On one well in 1997, I noticed the chemicals failed to create the normal viscous, jello-like fluid and they ended up pumping a much thinner fluid. The service company did not like it but we went ahead with the stimulation. The well turned out not all that bad and I thought I may be on to something.”
His idea got a boost when he met Mike Mayerhofer, an engineer working at Union Pacific Resources, who had been successfully using mostly water for fracking operations in East Texas, although they added polymers to reduce the water’s surface friction. This reminded Nick of his experience with the watery gel-frack, so to cut costs and hopefully increase production, the use of ‘slick water’ fracking was proposed to management.
“We got approval for the slick water fracks on three wells in spring 1997,” recalls George Jackson, then heading up Mitchell’s engineering effort. “Most of the folks involved were against it but we really needed to cut costs. The slick water fracks did not flow back very well at first. We were possibly overly cautious and restricted the wells on flowback thinking we could crush the resulting fractures. After a month, we opened up the wells again and they flowed gas at decent rates similar to the rates after a gel frack.”
“During the winter of 1997–98, we were able to stimulate additional wells,” says Nick. “However, instead of the Union Pacific Resources formula, we increased the frack size using greater volumes of water but lower concentrations of sand (now called light sand fracks) to create more micro-fractures. When the production from the S. H. Griffin No. 4 well was doing better than any well had ever done, I thought we had found the key.” The production from this well was 1.3 MMcfpd gas after its first 90 days. No Barnett well had ever done that – and it kept going. Steinsberger, in an interview with The Atlantic, said, “This was the ‘aha moment’ for us, it was our best well ever in the Barnett, and it was a slick water frack. And it was my baby.” Not only did slick water fracks work better, they only cost $85,000, saving nearly $300,000 per completion. George Mitchell was well on his way to his goal to replacing his gas source.
Yet There Was More!
Enter a new geologist: Kent Bowker had been working for Chevron on its unconventional gas acreage, south of where Mitchell was working, and drilled a Barnett test in 1997: a 60-foot (18m) conventional core taken in the middle of the formation. “We put the cores in rubber sleeves and I remember seeing the ends bulging with gas after a short time,” says Kent. “My thoughts right then were that this formation had a whole lot more gas than previous studies indicated.” Indeed, earlier absorption isotherm tests performed by GRI contractors were much lower than the later tests by Chevron and Mitchell. “The well proved to be a failure for Chevron,” says Kent. “However, the results of the higher gas content convinced Mitchell to reevaluate the gas content of its Newark East field.
“By 1999, with the higher gas content and much reduced completion costs, the Mitchell team started to complete the Upper Barnett and re-fracks on existing wells,” Kent continues. “The Upper Barnett added an average of 250 MMcfg reserves per well and the re-stimulated wells that had been gel-fracked were adding 500 MMcfg additional reserves on an average per well basis. (The new stimulations created a much bigger fracture network than the older gel stimulations.) What makes the Barnett so prolific is the huge amount of gas in place, in an overpressured and fully-saturated state.”
The George Mitchell Factor
While George Mitchell did not invent fracking, his interest in the Barnett and tight gas plays went back to the 1950s when he noticed a layer of impermeable rock in wells drilled north of Fort Worth, Texas. He thought this could contain a lot of natural gas and obtained leases in the area, drilling his first well there in 1951. The rock was tight sandstone with natural fractures. Many subsequent wells there were hydraulic fractured, a technique developed commercially only a few years earlier.
Mitchell kept his eye on the Barnett and prompted his team to test it in 1981. “George was a strong-minded, stay on top of it type of person,” says Loren Steffy, Houston author and biographer for George Mitchell (book currently in review). “For the Barnett, he brought perseverance to the equation. He spent millions more than any other company over the 17 years it took for success. Once he was convinced he had a gas source, it did not matter how hard or long, he was going to see it through.”
George Mitchell controlled about 70% of the company, allowing him to override his board and even his own engineers. Jon Huggins remembers, “He was difficult to work for at times and had his own ideas. Initially, engineers were sceptical of pursuing the Barnett, but over time became believers. The H. A. Smith well in Denton County, east of established production, was plugged when logs showed it did not meet the net porosity cut-off that had been established for the play. When George became aware of this, he made the decision to unplug and complete the well. He kept the group focused on the mission and on trying until we got it right.”
Mitchell went on to prove up more acreage prior to selling the company to Devon, who with other companies took the lead in developing the field using horizontal drilling and multi-stage stimulations. Newark is now listed as the second largest gas field in the US, only surpassed by the Marcellus of Pennsylvania and West Virginia, another shale play that was developed initially using knowledge learned in the Barnett.