This map shows the location of Mitchell Energy’s first Barnett Shale well. (Images courtesy of Republic Energy)

Mitchell Energy began the Barnett Shale play in North Texas with the drilling of the C.W. Slay #1 in 1981 in southeastern Wise County (Figure 1). Even though this initial test was productive, it was not overly encouraging. Initially, the exploration model was dependent on open natural fracture networks as defined in the Appalachian Basin Devonian gas shale play (then the only shale model). The Slay #1 had only been given a small frac, and the assumption was made that there was very little natural fracture network associated with the well.

During the 1980s Mitchell’s Barnett team deepened Boonsville Bend Conglomerate gas field wells in a four-county area to build a database, develop structural and isopach maps, and evaluate organic richness and thermal maturity of the rock. Once the wells were drilled and cased, various completion and stimulation techniques were evaluated. Through 1986, 41 wells were deepened to the Barnett and many were given massive hydraulic fracture treatments varying in size from 250 to 1,500 ft (76.3 to 458 m) theoretical frac half lengths (TFHL). By 1988 sufficient well and seismic data were obtained to begin understanding the deep structure in Wise County. In addition, an area in the county nine miles northwest of the Slay #1 had been identified where a step-wise increase in frac size yielded a corresponding increase in initial flow rates with repeatable results (Figure 2). Using 1,500-ft TFHL, initial flow rates of 0.9 to 1.2 MMcf/d were achieved. However, this initial area amounted to only 15 sq miles (39 sq km).

Initial development

The company began development in this area in the late ’80s and continued to probe outward in all directions. Modifying the frac allowed the enlargement of what was considered the core area. This growth encompassed an area in Wise County bounded on the west by the Viola-Simpson erosional limit, on the north by the Mineral Wells fault zone, on the south by the Rhome-Newark fault/Forestburg limestone pinchout and on the east into a portion of southwest Denton County. This new area covered approximately 115 sq miles (298 sq km) through 1997 (Figure 2). By 1992 results of pre-frac testing, multiple Barnett cores and post-frac decline curve analysis indicated that the play was not dependent on open natural fractures as originally thought but required large induced fractures to establish commerciality. The size of the induced fractures was dependent on productive section thickness and frac cost. Using a THFL of 1,500 ft appeared to optimize initial flow rates and estimated ultimate recoveries (EUR) to well economics.

In 1997 light sand (slick water) fracs in the core area were being tested in an attempt to improve economics after a decrease in realized gas prices. By October 1998 these new fracs were found not only to reduce stimulation cost by approximately 50%, but in most cases they also resulted in similar initial rates and higher subsequent stabilized rates. Outside the core area the team’s initial theory was that the absence of the Forestburg limestone above the lower Barnett or the absence of Viola-Simpson limestone below had resulted in the loss of the reservoir seal, which was assumed to account for poor completions in these areas. However, by late 1998 it was determined that areas with thin or absent Forestburg limestone were poor performers not because of lack of seal but due to greater frac height caused by insufficient upper frac containment. In early 1999 Mitchell’s production engineers began testing areas without Forestburg by perforating both the upper and lower Barnett and doubling the size of the light sand frac. This proved very successful, and the core area was expanded south of the Rhome-Newark fault zone, taking in the area of the C.W. Slay #1 discovery well and going south into Tarrant County. Light sand water fracs were then applied north of the Mineral Wells fault zone and west of the Viola limestone pinchout, proving successful in some areas but not everywhere. After continued testing it became apparent that wells located in the gas-prone area which were not commercial resulted from poor frac efficiency, not poor Barnett. Application of microseismic frac mapping technology provided some of the support for this conclusion as well as Mitchell’s preliminary horizontal testing in a non-core area in 1998. Mitchell was successful in drilling two horizontal wells beyond the Viola pinchout, and both were fractured with fluid volumes comparable to vertical wells in the core area to the east. Although both wells were mechanical successes in that the fracture treatments stayed in zone, they were considered commercial failures since the resultant rates and EURs would not recoup total cost of the well.

Mitchell engineers believed commerciality could be achieved by increasing the size of the frac. However, when the company was offered for sale in late 1999, additional horizontals were put on hold. From early 2000 until May 2001, when Mitchell drilled its 600th Barnett well, the company focused on drilling development wells, improving the gathering system and expanding the gas plant for the purpose of increasing production and industry awareness. This proved successful as Mitchell Energy and Devon Energy finalized their merger in January 2002.

Within four months of the merger Devon permitted its initial horizontal wells; the first of these wells came on production in December 2002 and January 2003. Two of these were drilled in non-commercial areas of Wise County: C.J. Harrison A-2 was west of the Viola Erosional Limit, and O.H. McAlister 16 was in an area of weak or fractured Viola limestone. Both were exceptional successes, using frac sizes 2-1/2 to 3 times that used for conventional vertical wells. The Harrison A-2 came on production at 2.5 MMcf/d, and the O.H. McAlister #16 came on in excess of 3 MMcf/d.

While massive gel fracs were responsible for the initial commercial phase of the Barnett, and light sand water fracs caused a rapid expansion of drilling in the play as well as enlargement of the play area, horizontals propelled the play to the status it has today. Horizontals, however, generally require an understanding of the structural complexities and dips of the area into which the lateral is being drilled. In late 2003 Republic Energy Inc. drilled the first successful horizontal in Parker County using 3-D seismic to select the location and direction of the well. Many attribute this successful use of seismic to the current insistence by most operators to have 3-D coverage before drilling horizontal wells in the play.

Drilling activity has expanded to 20 counties in the North Texas area, primarily due to horizontal drilling (Figure 3). Mitchell Energy’s experience in the early years and industry’s knowledge in general in the later years have shown that repeatable success in the play is primarily dependent on two things: being in the gas-prone area of the Barnett where paleo temperatures have cracked the oil to gas; and development of technology for individual areas regardless of Barnett thickness, structural complexity or current depth of the section.

The current state of knowledge places the western limit of the gas prone area as cutting southwest diagonally across Jack County south into western Palo Pinto or eastern Stephens Counties, then south along the west line of Erath County and southeast through the eastern quarter of Hamilton County. Operators are working on technology to stimulate the thinner Barnett sections found to the west and have used gas lift systems to improve after-frac clean-up in the shallow areas. The Barnett is an unconventional tight gas technology-driven play, and its success can be attributed to patience, perseverance and the commitment of many within the industry. While the play is important with respect to its gas reserves, its real value comes more from the understanding and technological development that has or will be achieved to adequately exploit source rock resource plays on a worldwide scale.

Editor’s note: Dan Steward authored The Barnett Shale Play, Phoenix of the Fort Worth Basin, a History. This book was published in spring 2007 by the Fort Worth Geological Society and North Texas Geological Society and is available for purchase through either organization.