THE BARNETT SHALE: NOT SO SIMPLE AFTER ALL
By Natalie Givens and Hank Zhao

Many special thanks to Dan Steward, Nick Steinsberger, Kent Bowker and Brad Curtis.

The technological advances made within the last decade have allowed the Barnett Shale (Barnett) to grow from a crazy prospect to being one of the largest onshore natural gas plays within the continental U.S.  With an estimated 26.2 TCF gas in place, the Barnett has begun to attract worldwide attention (USGS 2004).  A Barnett well in the heart of the Newark East Field has an average depth of 7,500', an Estimated Ultimate Recovery (EUR) of 1.25 BCF, the possibility of multiple fracture stimulations (refracs) and a cost of roughly  $600,000 per vertical well.   But don’t get too excited yet...there are many hurdles we must understand to get good Barnett production.

We must know and try to understand several items of importance: 

• The maturation pattern of the Barnett in North Texas
• Regional faulting and underlying Ellenberger karsting
• Fracture stimulation (frac) techniques designed to meet the needs of a given area
• The thickness of the Barnett across the prospective area
• A drilling and completion strategy so as not to inundate the Barnett with frac water

Each of these factors has a significant effect on ultimate reserves and must be appreciated with respect to each other. Yes, the Barnett, in varying thickness, is always present in the Fort Worth Basin.  What might seem simple to those familiar with structurally and stratigraphically complex plays, is not so simple after all.  Over the past two plus decades many geoscientists and engineers have gone to great efforts to understand the geology and geochemistry and their impact on successful drilling and completion techniques for this prolific formation.

Figure 1: The extent of the Fort Worth Basin, bounded on the East by the Ouachita Thrust, on the North by the Muenster Arch, on the West by the Bend Arch and on the South by the Llano Uplift.   Also shown are the Gas Maturation Limit, the Newark East Barnett Shale play Expansion Area, the structure at the base of Barnett Shale, the Viola Erosional Limit, the approximate boundary of the Fort Worth Basin, and the cross section line A-A’ for Figure 9.  Modified from H. Zhao.

Geologic Background

The Barnett Shale was deposited over present day North Central Texas during the Late Mississippian Age in a time of marine transgression caused by the closing of the Iapetus Ocean Basin (Henry).  By the end of the Pennsylvanian the Ouachita Thrust belt began encroaching into the present day North Texas area.  The thrust belt owes its existence to the subduction of the South American plate under the North American plate.  The Ouachita Thrust's emergence created the foreland basin along the front of the thrust. Early studies of the basin attributed thermal maturation of the Barnett to burial history and the thermal regimes associated with depth of burial.  Explorationists began to doubt  this hypothesis as more data became available.  Kent Bowker formerly of Mitchell Energy/Devon proposed a different model (Bowker) suggesting the maturation process  was driven by displacement of hot fluids, from east to west, associated with the Ouachita Thrust.  Figure 1 shows the extent of the Barnett in North Texas.

Maturation

Due to the lack of available Vitrinite Reflectance (VRo) geochem data Barnett BTU data is being used by many in the play to define the gas prone areas of the Barnett.  BTU values are still being examined within industry, but generally speaking values of 1380 BTU or less are thought to be associated with gas prone Barnett and values above 1400 BTU are thought to be in oil prone areas.

Tarrant, Johnson, Hood, Somervell, Bosque, most of Wise and Parker and portions of Dallas, Denton and Ellis Counties are interpreted to be in the gas window; whereas Clay, Montague, Cooke and most of Jack Counties are in the oil window.  The western extent of the gas prone area has yet to be determined.  Figure 2 shows contours outlining areas of numerous BTU values.  These contours are based on various well data from different operators across the area.

Analysis of conventional and rotary cores proved to be beneficial in understanding rock properties as well as production characteristics and geochem data.  Mitchell Energy had reportedly taken in excess of 1,000' of core throughout the productive area for the purpose of equilibrating logs to rocks.  Core taken from Mitchell's W.C. Bill Young #2 in 1983, a Wise County test well, showed that the permeability was very low, from 0.001 to 0.00009 millidarcies (md), the porosities averaged approximately 3.5%, the calculated water saturation was in the mid 40% range and 75-80% of all natural fractures were healed.  Composition of the Barnett is 2-8% organics, 20-30% Clay minerals (Illite), 45-55% Silt (Quartz and Feldspar), 15-19% Carbonates (Calcite and Dolomite) and trace amounts of Siderite and Pyrite.  Additional core was taken in the early 1990's in conjunction with a GRI project which supported this earlier analysis.  The Barnett is the source rock, the reservoir rock and its own seal.

Figure 2: The Expansion area and contours of BTU data derived from various operators’ wells throughout the area and an approximate Gas Maturation Limit.

Field History

Mitchell Energy had drilled in excess of 3,500 wells in the Fort Worth Basin during the 50 year period prior to its sale in January 2002 to Devon Energy.  Until the 1980's most of this drilling had been for the prolific Boonsville Bend Conglomerate and Strawn reservoirs that dominated the Basin's production.  Mitchell's first well completed in the Barnett Shale was the C.W. Slay #1 in south-eastern Wise County, TX.  The Slay #1 was an exploratory deepening of a Rhome Caddo development well drilled in 1981 to evaluate Viola Limestone potential; however, the Viola was tight.  The log similarities in the Slay #1 between the Barnett Shale and the Devonian Shales in the Appalachian Basin were apparent.  Knowing the prolific nature of the Appalachian Devonian Shales the geologic and engineering teams at Mitchell began to devise a plan to deepen more wells to create a Barnett database.  From 1981 to 1989 there were only 63 wells drilled, mostly by Mitchell Energy.  These wells were used to develop a better understanding of the formations producibility and develop completion techniques.  The early stimulations used in these wells reportedly progressed from small CO2 or N2 fracture treatments to gel fracs.  Mitchell adopted a massive hydraulic gel frac (MHF) in 1986 that had a theoretical half frac length of 1,500'.  A typical design would have had approximately 400,000 gallons of water and 1,250,000 pounds of sand, (Fig. 3 and 4).

Figure 3:  Production for the C.W. Slay #1.

From 1990 through 1994 an additional 200 wells were drilled including the first horizontal well in the Newark East Barnett Shale Field, the T.P. Sims B #1 in Wise County, TX.  By this time, enough production had been obtained to assemble a useful decline curve and an appreciation of per well reserves could be recognized.  In addition, developmental areas were identified where these decline models would work.  By the mid 1990's Mitchell was able to assign 1.0 BCFG EUR per well in these development areas.   From 1995 through 1999, 365 wells were drilled.  During the early part of 1997 Mitchell Energy tried a different technique to stimulate the Barnett.  The first slick water frac, also called light sand frac (LSF), was scheduled.  The LSF was very successful at stimulating the Barnett and decreasing the completion costs of the wells.  The early LSF's had 800,000 gallons of water and 200,000 pounds of sand.  Micro-Seismic fracture mapping from many of these wells indicated that the lower viscosity water was invading partially healed natural fractures and subsequently breaking them down.  This evidence was proved when wells within 1,500'-2,000' of each other in any direction were being affected by the water pumped in the well being treated with a LSF;  Whereas, the gel fracs would only affect other wells if they were very close together and in a NE-SW orientation from one another.  Micro-Seismic also confirmed earlier work indicating that the general trend of the induced fractures was NE-SW.  Figure 5 shows an example of a map view from Micro-Seismic frac mapping.  The extension and placement of the induced fractures is easily visible in this format.  Frac containment within the shale is visible in Figure 6.

The first refrac of the Lower Barnett occurred early in 1998.  Mitchell's B.S. Carter Jr. #4, was drilled and completed in 1995 and was treated a second time with a LSF (Fig. 7).  Production curves have demonstrated that refracs of primary LSF wells have enhanced production, but to a lesser degree than a LSF refrac of a primary MHF well, initial tests were performed in 2001.  It is thought that for LSF refracs of a primary LSF frac job to be optimized, cumulative production before fracing must be greater in a LSF well than in a refrac of a primary MHF well.  This can be attributed to the greater amount of frac area exposed in a LSF and a smaller amount of gas depletion per square foot for a given cumulative.

Figure 4: Timeline of wells drilled and major fracture technique breakthroughs.

Figure 5: Seismic data from a horizontal well in map view.  Visible are dots representing strong seismic events induced by the fracture stimulation, these are measured by the geophones down hole in the observation well.  The red lines represent possible connectivity of induced fractures.  Red dots represent the first fracture stage and blue the second

Figure 6: The same seismic events shown in Figure 5, but in cross section.  The events show that containment within the Barnett was successful. 

Figure 7: Production for the B.S. Carter Jr. #4.

Figure 8: The Core area for Barnett production and the Counties included in the expansion of the play, the Viola Erosional Limit, the Forestburg Limit, and the areas of Single, Double and Multiple Stage Fracture Stimulations.

Another saving grace for the Barnett occurred between fall 1998 and spring 1999 when Mitchell started Another saving grace for the Barnett occurred between fall 1998 and spring 1999 when Mitchell started testing the Upper Barnett's productive capabilities.  Reserves for the Upper Barnett were estimated at a quarter of a BCF.  Shortly thereafter Mitchell appeared to adopt a standard operating procedure of including the Upper Barnett in all its Barnett wells, thereby increasing the EUR per well from 1 BCF to 1.25 BCF.

The year 2000 had 186 wells drilled. The play was really starting to heat up by the end of 2000.  Gas prices had risen to an unforeseen and unbelievable $9.00+, Landsmen were tripping over each other at the court houses, drilling rigs were brought in by the dozen, like in other parts of the country during this gas boom, wells could not be drilled fast enough.  As a result 520 wells were drilled in 2001.  Mitchell had between 8 and 18 drilling rigs running to keep pace with their 120,000 acres in the core area of the field throughout these busy years (Fig. 4 and 8).

The 473 wells drilled in 2002 are a clear sign that the frenzy was not over.  Devon continued drilling vertical wells and also started their horizontal program drilling 9 horizontal wells in 2002.  As Devon engineers and geoscientists perfected their techniques for horizontal wells, the rest of the operators in the play waited for results with baited breath.  Finally, word leaked out, as it always does, that horizontals could be wildly successful in the core area and would probably lead to the expansion of the play (Fig. 8).  Still most of industry waited for solid proof.  Production history of these wells was slow in materializing through the Railroad Commission of Texas.  Other operators began their own experiments with horizontals in the core area during 2003.  Of the 780+ wells drilled in 2003, roughly 130 were horizontal, drilled by 27 different operators.  A successful horizontal well bore not only increased the rates of these wells by at least two times, it also slowed the decline curve and allowed expansion of the play to areas where there are no upper or lower frac containment barriers.  However, a firm EUR value for horizontal wells in the Barnett has yet to be decided upon.  For a horizontal well to be highly successful the frac must be contained within the Upper and Lower Barnett Shale.  The core area as an upper frac containment barrier in the Marble Falls Limestone and a lower frac containment barrier in the Viola Limestone.  The Viola Limestone erodes (pinches out) to the west and southwest of the core area.  Vertical wells have been tried in the areas to the west of the Viola erosional limit, (Fig. 9), but overall have not been very successful.  Many have frac'd into the Ellenburger thereby reducing the frac efficiency and opening a conduit for water production.  Horizontal technology at this time appears to be the most effective avenue of expanding the play into previously noncommercial gas prone areas.

Figure 9: Schematic of the lithology in the Fort Worth Basin.

Regional Faulting and Karsting

Mitchell Explorationists had found that wells drilled on a fault of any kind tend to have high frac gradients, this also proved to be true with regard to wells drilled on prominent structural highs.  In addition, wells drilled near major tectonic faults tended to frac toward the fault and break down the fault plane (Fig. 10).  This generally resulted in low frac efficiency and communication with Viola or Ellenburger water.  Mitchell had been applying geophysics to the area since the early 1980's and in the early 1990's shot the first 3-D in the basin.  By 2002 they had been involved in over 250 sq. miles of 3-D, with most of it in the Barnett gas prone areas.  Seismic surveys demonstrated the highly karsted nature of the Ellenburger and its effect on the overlying section, creating failure chimneys up through the Barnett and frequently as high as the Caddo Limestone in the Atoka section.  At this stage of development the faults associated with the karsts are considered an unnecessary hazard to Barnett completions (Fig. 11).  Figure 11 shows an area that is extremely karsted.  The largest karst in this area is roughly 200' deep and covers an area of approximately 200 acres.  Through the use of seismic, these karst features can be avoided and when that is not possible, then at a minimum the operator can cement the production string and not perforate near the faults. Regional Faulting and Karsting.

Figure 10: Large regional faults (blue lines), the arial extent of possible Ellenburger Karsting and the A-A' line of cross section for Figure 9.

Fracture Stimulations

As we can see from Figure 12, the Barnett thins towards the west and south away from the MuensterArch.   The northern portion of the core area has a 100'-300' thick Forestburg Limestone between the Upper and Lower Barnett.  Areas with a thick Forestburg Limestone generally have a two stage frac, one stage in the Upper and one in the Lower, to effectively stimulate both Barnett zones.  Operators typically perf and frac the Lower Barnett, set a bridge plug in the Forestburg Limestone, and then perf and frac the Upper Barnett. The southern portion of the core area has little to no Forestburg Limestone and thus only needs a single stage frac (Fig. 8 and 13).  The Lower Barnett can be divided into five shale zones due to the 10'-30' limestone intervals.  In the area approaching the Muenster Arch, the east side of the core area, these limestone intervals greatly increase in thickness, to 80'-150' thick.  These limestone intervals add to the gross thickness of the Barnett, but do not contribute much to the net thickness.  In the areas closest to the Muenster Arch three to four stage fracs are being utilized to maximize frac length within each zone. Fracture

Figure 11: An area of 3-D seismic coverage featuring extensive karst formation in the Ellenburger.  Structure on top of the Ellenburger.

Drilling and Completion Strategy

Operators have learned the hard way that it is important not to drill and complete wells in close proximity to each other too quickly.  When wells are stimulated using nearly a million gallons of water, it is easy to imagine that even with a good 2-3 day flow back of frac water, there is still a lot of water remaining in the formation.  When drilling is done on 40 acre spacing and fracture stimulations are only weeks apart, the induced fracture networks appear to be more limited on a per well basis, due to a lack of depletion around the well bore and to the water left in the induced fractures from nearby wells, causing clean up to be more difficult.  The fractures need time to deplete an area to a point where subsequent fracs will seek out new rock rather than following existing patterns.  While industry believes that vertical wells are draining less than 40 acres, it is best not to infill this pattern early in the development.  Infill wells should be drilled after frac water has been produced and some depletion has occurred around the fractures.

Figure 12:  Isopach of the Barnett Shale, 25' Contour Interval.  Irregular contours in NW Johnson County due to karsting.

Figure 13: Type log for southern core area.

Conclusions

The evolution of the Newark East Barnett Shale Field can be attributed to many talented individuals.  The Newark East Field is a technology driven play.  Industry has had to challenge and adapt techniques to fit the Barnett.  Through the use of geochem, cores, seismic, micro-seismic and many other tools, operators have been able to unravel some of the Barnett's secrets and make the play successful.  The Barnett has challenged many of our conventional concepts both from a geological and engineering standpoint.  The industry has been forced to not only look at things from totally new perspectives, but in some cases completely throw out old concepts.  Application of state of the art technology is one of the most significant aspects of this play.  Pushing the technology envelope to not only understand the rock, but to optimize production from one area to the next. 

Works Cited and References

Angevine, C.L. and Turcotte, D.L.  Oil generation in Overthrust Belts.  AAPG Bulletin, V. 67, No.2 (February 1983), pp. 234-241.

Bowker, Kent.  Recent Development of the Barnett Shale Play, Forth Worth Basin. 

West Texas Geological Society Bulletin, February 2003. Volume 42, Number 6.

Curtis, John.  Fractured shale gas systems. AAPG Bulletin, V. 86, No. 11 (November 2002),  pp. 1921-1938.

Henry, James D.  Stratigraphy of the Barnett Shale (Mississippian) and Associated Reefs in the NorthernFort Worth Basin.  Petroleum Geology of the Forth Worth Basin and Bend Arch Area, pp. 157-177.  Charles A Martin Editor.  Dallas Geological Society, 1982.

Jarvie, Dan.  The Barnett Shale as a Model for Unconventional Shale Gas Exploration. Website slide set, www.humble-inc.com, presented at the AAPG Southwest Section Meeting, June 2002, Ruidoso, New Mexico.

Perry, William J., Jr.  Structural Settings of Deep Natural Gas Accumulations in the Conterminous United States.  USGS Bulletin 2146-D, 1997, pp. 41-46.

USGS. Assessment of Undiscovered Oil and Gas Resources of the Bend Arch-Forth Worth Basin Province of North Central Texas and Southwestern Oklahoma, 2003.  Fact Sheet 204-3022, March 2004.