October 2000
Special Focus

Horizontal drilling proves cost-effective in boosting gas production

Texaco describes evolution of the technology that made drilling of 50 horizontal laterals in a Permian basin gas field an economic success

Oct. 2000 Vol. 221 No. 10 
Feature Article 


Horizontal drilling proves cost-effective in boosting gas production

A 30-fold increase in gas production resulted from an evolving horizontal drilling program that ultimately cut drilling time by 40%

Glenn W. Cox, Texaco E&P Inc.

The recent development of Bryant-G-Devonian field has been a significant success for Texaco in the Permian basin. The field had been producing about 2 MMcfgd with 100 bpd of associated condensate. Through the application of enhanced reservoir imaging and horizontal drilling technology, production was increased to about 60 MMcfgd and 2,600 bcpd, with expected recoverable reserves being increased by 300%. An additional benefit of this program was an increase in loading at a Texaco-operated gas plant, which processes liquids from this gas stream.

During this ongoing program, about 50 horizontal laterals have been drilled, either as re-entries from existing wells or from "grass roots" new wells – some with single laterals and some with multiple laterals. Since this was the largest package of deeper (>10,000 ft) horizontal wells Texaco had developed in the Permian basin, the learning curve was relatively steep and oftentimes bumpy. The discussion below outlines some of the progressions followed to bring drilling practices to their current form.

Field History

Bryant-G-Devonian field (BGDF) is located in western Midland County, Texas, in the Midland basin region of the Permian basin. The field was originally discovered and developed in the mid-1960s with 19 wells drilled in eight sections. These original wells were multiple completions in the Pennsylvanian and Wolfcamp formations, as well as the Devonian. A rapid decline in oil:gas ratio indicated that hydrocarbons existed as a retrograde condensate reservoir. In light of the success seen at the near-by Headlee Devonian Unit, the Devonian formation was unitized as the Bryant-G-Devonian Unit (BGDU), and a gas cycling project was initiated in the mid-1970s.

Although a moderate condensate production increase of 300 to 400 bcpd was achieved, reservoir permeability proved to be too low and too discontinuous to allow effective sweeping between the injection and production wells. As a result, the gas cycling project was deemed an economic failure and gas re-injection was soon abandoned. After returning to primary production, the field continued to decline until stabilizing with virtually no decline at about 2 MMcfgd and 100 bcpd. This production profile is again consistent with inadequate drainage of a low permeability reservoir.

To increase field recovery, an infill drilling program was carried out in 1994 and 1995. Eleven additional wells were drilled in the field, resulting in good initial production rates but rapid declines, with the wells stabilizing at around 500 Mcfd per well. The vertical wells in the field (Fig. 1) were completed with 5-1/2-in. casing and stimulated with acid and/or sand fracs. During this time, Mobil drilled and reported good success from the first horizontal well in the adjacent Parks (Penn) field.

Fig 1

Fig. 1. Locations of vertical wells in Bryant-G-Devonian Unit are shown. Eleven additional wells were drilled in 1994/1995, resulting in good initial production rates but rapid declines. Wells stabilized at about 500 Mcfd per well.

Since Texaco was involved in an initiative to apply horizontal drilling technology in applicable reservoirs and fields, the decision was made to re-enter an existing wellbore to drill horizontally in the BGDF. By transferring technology used successfully at the Aneth Unit in southeast Utah and at Sundown Slaughter Unit near Levelland, Texas, the BGDU 23-H was re-entered and drilled to a lateral length of 1,566 ft during March 1996.

After acid stimulation, the well was potential tested at 4 MMcfd and 100 bcpd. This success resulted in a ramped up drilling program during 1996 and 1997, consisting of 11 re-entries and 19 new wells within the unit. In addition, 13 new wells tested the flanks of the structure outside of the unit boundary. After a slowdown during 1998, the program in 1999 concentrated on adding second laterals to existing wellbores. The current field map is shown in Fig. 2.

Fig 2

Fig. 2. Current field map shows results of ramped up 1996/1997 drilling program, which consisted of 11 re-entries and 19 new wells. In addition, 13 new wells tested the structure’s flanks outside the unit boundary. The 1999 program concentrated on adding second laterals to existing wellbores.

Geological Summary

BGDF produces from the Thirtyone formation (known locally as the Devonian). In central Midland County, the Thirtyone comprises a 600- to 700-ft shoaling-upward sequence of ramp carbonate debris, which was derived from a prograding carbonate platform to the north. The carbonate sediment ranges from a lower-ramp, chert-rich, carbonate wackstone (fine-grained) at its base, to a coarse-grained, upper-ramp, crinoidal grainstone toward the top. Both the wackstones and crinoidal grainstones lack porosity and permeability.

Most of the reservoir quality rocks at BGDF are found in sequences of packstones and grainstones, rich in siliceous sponge skeletal remains known as spicules. Many of these spicules have been locally dissolved leaving a network of needle-like pores. Although the dissolution of the spicules enhance reservoir porosity, much of the dissolved silica was re-precipitated above and below the reservoir layers as nodular chert. These nodular chert-rich layers can play havoc on drill bits. Individual reservoir units range in thickness from 5 to 40 ft.

The top of the Devonian occurs at about 11,700 ft, with the primary pay located 100 to 200 ft deeper. Immediately above the Devonian is the Woodford shale. This 70-ft-thick interval is extremely water sensitive and cannot be left exposed for long periods of time.

Re-Entry Procedures

The first horizontal tests at BGDF were drilled as re-entries from existing wellbores in order to utilize existing well log information and to avoid costs of a new wellbore. The 5-1/2-in. casing in the newer wells limited maximum hole size to 4-3/4 in., while the available distance between the base of the Woodford shale and the target porosity demanded the use of what was at that time considered "short-radius" drilling technology.

For the original test, the casing exit from the 5-1/2-in., 20-ppf casing was achieved by section milling from 11,745 ft to 11,785 ft. This section was then underreamed to 8-1/2 in. before setting a cement kick-off plug. After dressing the plug to a kick-off point of 11,750 ft, a 4-3/4-in. curve was drilled, building angle at 48° per 100 ft.

The bottomhole assembly for drilling the curve included a 4-3/4-in., IADC type 537 bit, bit sub, 3-3/4-in. downhole motor with a 3° bend and two articulations, float / orienting sub, two flexible non-magnetic drill collars, and 2,260 ft (sufficient for the expected lateral length) of L-80 tubing with PH-6 connections. The remainder of the drillstring was 2-7/8-in. drill pipe with American Open Hole connections.

After drilling the kick off with gyroscopic surveys, magnetic directional surveys were obtained with wireline steering tools. The curve was landed with an inclination of 80.5° at a measured depth of 11,898 ft and a TVD of 11,852 ft. The BHA was tripped for the lateral assembly in which the motor was exchanged for a 3-3/4-in. motor with 1.5° bend and a single articulation. This type assembly was used to drill the entire lateral to a final TD of 13,316 ft. Throughout this entire program, laterals have been completed openhole.

Ten wells that were drilled in 1994 and 1995 were re-entered over the course of the next eight months. Horizontal laterals up to 3,800 ft were successfully drilled using these techniques. As the program continued, several modifications and improvements were implemented to reduce overall drilling costs.

Whipstocks. Section milling was eliminated in favor of using whipstocks to mill windows through the casing. At these depths, whipstocks tend to provide a more reliable and cost-efficient casing exit than can be achieved with milled sections and kick-off plugs. Fig. 3 shows a schematic wellbore diagram of the typical re-entry horizontal.

Fig 3

Fig. 3. Schematic wellbore diagram shows a typical horizontal re-entry completion.

Measurement while drilling. MWD tools replaced steering tools as that technology adapted to the slimhole, short-radius equipment. This reduced the time consumed by tripping wireline for directional surveys.

Motors. When the program was started, conventional wisdom dictated that articulated motors were necessary in this hole size to achieve desired build rates. Application repeatedly showed that motors built more angle than predicted by the manuals. Ultimately, fixed-bend motors, without articulation, were used to successfully drill curves with build rates up to 48° per 100 ft. This change reduced the complexity and increased the reliability of the curve drilling assemblies.

Drill time curves for the first and the last re-entries drilled in the program are shown in Fig. 4. Although there are still potential difficulties involved with setting whipstocks and milling windows, as seen by the extended prep times, ultimate lateral section lengths were extended significantly.

Fig 4

Fig. 4. Drill time curves for first and last re-entries show that there are still potential difficulties involved with setting whipstocks and milling windows, as seen by the extended prep times. However, ultimate lateral section lengths were extended significantly.

New Drill Procedures

Concurrent with the re-entry program, additional infill wells were drilled to completely develop the field consistent with field rules and to test the extent of commercial reserves. A desired hole size of at least 6 in. for new laterals dictated 7-in. casing for the production string, hence 9-5/8-in. intermediate casing and 13-3/8-in. surface casing.

Surface casing was set at 350 ft, with cement circulated to surface to protect fresh water sands. Intermediate casing was then set at 5,000 ft to cover the porous section of the San Andres. It too was cemented to surface. To confirm and evaluate the pay, production holes were typically drilled through the Devonian to allow logging and possibly coring of the pay interval. Various methods to initiate the horizontal laterals were applied.

Cased curve. The first well designed specifically as a horizontal producer was planned to have casing set around a medium-radius (±500-ft) curve. A medium radius design was selected to assure that a 4,000-ft lateral could be achieved.

Well BGDU 31-H was drilled vertically to a kick-off point (KOP) of 11,436 ft before directional control commenced. Angle was built at about 12° per 100 ft to land the curve at 12,328 ft (11,925 ft TVD). Casing was run to TD and cemented. After drilling out cement, a 6-1/8-in. lateral was drilled to a measured depth of 16,308 ft. To better define the desired curve landing point on subsequent wells, pilot holes were drilled through the Devonian to about 12,200 ft. Open hole logs and side wall cores were obtained before the wellbores were plugged back to 11,400 ft. At that point, a medium-radius curve and lateral were drilled as described above. A wellbore schematic of this technique is shown in Fig. 5.

Fig 5

Fig. 5. Wellbore schematic illustrates cased curve technique. To better define desired curve landing point, pilot holes were drilled through the Devonian to about 12,200 ft. After logging, wellbores were plugged back and a medium-radius curve and lateral were drilled.

Top set. As seen on the drill time curve, a significant period of costly rig time was being consumed between reaching total depth on the vertical hole and drilling new hole on the lateral. Directional tools and personnel also had to be kept on standby while running casing and drilling out cement. Since success had been seen drilling laterals in excess of 3,500 ft in the re-entry program, a method of drilling short-radius curves below the Woodford shale was developed.

For BGDU 34, the 8-3/4-in. production hole was drilled to 11,750 ft and 7-in. casing was set in the top of the Devonian limestone. After drilling out cement, whole core was taken and a pilot hole was drilled to 12,200 ft. Openhole logs were run at TD and the well was plugged back to a KOP of 11,785 ft. Directional tools were picked up and a 6-1/8-in. curve was drilled and landed at 11,970 ft (11,917 ft TVD). The lateral was then drilled to a measured depth of 15,836 ft.

A variation of this method was to drill 8-3/4-in. hole through the Devonian and log the entire hole. Then, the hole was plugged back and casing run to the top of the Devonian. After drilling out float equipment, directional activity commenced. This technique is shown in Fig. 6.

Fig 6

Fig. 6. The top-set technique is a variation of the cased-curve method. An 8-3/4-in. hole is drilled through the Devonian, entire hole is logged and the hole is plugged back and casing run to the top of the Devonian. After drilling out float equipment, directional activity commences.

Case-through / whipstock. Although top-setting casing in the Devonian successfully reduced the turn-around time and cost between drilling the vertical portion of the hole and drilling the lateral, there were still delays associated with obtaining competent kick-off plugs. There was also a need to more fully evaluate log information prior to selecting target depths and kick-off points. On subsequent wells, the decision was made to set casing to TD after logging and to utilize whipstocks to initiate drilling the curves. An additional benefit to the whipstock method of casing exit is that, if KOPs are selected properly, multiple horizontal laterals can be drilled from a single wellbore.

Consistently prolific production rates from the first few wells demanded that the program be accelerated on a two-year schedule. Four drilling rigs were assigned to the project, with two rigs drilling and setting casing on vertical holes, while the other two rigs cut windows and drilled the laterals. This method allowed use of a service unit to drill out cement and stage tools and to set whipstocks prior to moving the lateral rig on the wells. The lower-cost service units offset the incremental cost of additional casing and whipstocks required for the case-through completions.

Mud Programs

During early stages of the BGDF horizontal project, it was reasoned that a mud system similar to that used to drill vertically through the Devonian would be required. That system consisted of an 8.5- to 8.7-ppg, fresh-gel mud with 35- to 50-sec funnel viscosity and a water loss around 5 cc. Conventional thinking was that the viscous fluid would be necessary to transport cuttings out of the lateral, while the low water loss would minimize filtrate invasion of the formation. However, with an average reservoir pressure equivalent to a 6-ppg gradient, the primary result from this mud system was the need to add large amounts of lost circulation material to the system with frequent differential sticking events and subsequent fishing jobs.

Beginning in late 1996, most laterals were drilled with clear, fresh water as the circulating medium, circulating through the reserve pit for solids control. While minor seepage losses were still seen, sticking problems were all but eliminated. System additives were limited to pH control and drillstring lubricants with occasional polymer sweeps for hole cleaning. Improved performance while sliding and reduced torque and drag were obtained with the addition of solid, spherical, plastic beads. Since circulation was through the reserve pit, a shaker-type, bead-recovery system was installed to minimize consumption of the beads.

Completion and Stimulation

As mentioned earlier, all laterals in this program have been completed openhole. Reservoir simulations run early in the program indicated that the wells should have been capable of flowing commercial quantities of gas without stimulation. A skin factor on the order of positive 200 was imposed on the face of the laterals before the model predicted noncommercial rates. However, the first laterals drilled in the field would not produce naturally and did require stimulation.

Early stimulation methods called for running tubing to near the end of the laterals and pumping 60 gal of 28% HCl acid per ft of lateral length in four to twelve stages. Equivalent amounts of gelled water were pumped in alternating stages in hopes of extending or diverting acid penetration. Fluids were pumped simultaneously down the tubing and down the casing / tubing annulus. Radioactive tracer surveys indicated that the toe and heel of the laterals received most of the acid, while the middle portion of the laterals received little-to-no stimulation.

To increase the acid coverage of the laterals, a patented method was jointly developed by Schlumberger (Dowell) and Texaco to divert acid through a series of ports placed in the treatment string. The ports are sized to regulate the flow of acid through each port with pressure drop. Tracer surveys indicated this method greatly improved the acid distribution along the length of the lateral.

In an attempt to optimize stimulation treatments in the field, acid volumes have been varied between 35 and 65 gal per ft. Acid concentrations have been varied from 28% to 15%. Injection rates range from 20 to 100 bpm. Emulsified acids, such as the SXE acid system also have been tested.

None of these variables, nor lateral length, nor lateral diameter, appear to have any reasonable correlation with initial production rates or ultimate expected recovery. The only variable that seems to have any impact on production is the well’s location in the field. While research into stimulation methods continues, current practice is to attempt to balance the relative costs of injection rate, acid concentration and total fluid volume for a treatment consistent with the expected quality of the reservoir.


The drill time curves in Fig. 7 show one of the earliest wells in the program, one well from the middle of the package and one of the last wells drilled. The 40% reduction in total rig days represents savings on the order of $550,000. Efficiencies developed over the course of the program in completion methods resulted in additional per well savings exceeding $100,000.

Fig 7

Fig. 7. Between the earliest and latest wells in the program, drilling rig days were reduced 40%, for savings on the order of $550,000. Additional per-well savings are in excess of $100,000.

No single factor can account for this magnitude of cost reduction. Rather, a continuous effort by everyone involved moved the project further along the learning curve. Established practices and paradigms were challenged and improvements were shared not only within the project, but also with other projects worldwide. WO


The author thanks Texaco management for permitting the publication of this article, Dana Rowan for providing the geologic field description and Al Blessen for engineering the early years of this program.

Connect with World Oil
Connect with World Oil, the upstream industry's most trusted source of forecast data, industry trends, and insights into operational and technological advances.