Completion/stimulation practices in Cotton Valley gas wells - Cotton Valley Reef Trend of East Texas

Potential of the prolific Cotton Valley Reef Trend in East Texas has not been realized due to ineffective fracture treatments. Studies indicate better reservoir definition and more selective stimulations can improve success rates

The potential for prolific gas production from expensive, deep wells in the Cotton Valley Reef Trend of East Texas is high. Eighty wells have been drilled to date, with 41 commercial strikes defining an outstanding exploration target. However, efforts to improve economic returns by applying large hydraulic fracture treatments to the low-permeability reservoirs have been largely unsuccessful.

Improved practices offer possible completion alternatives. The following article develops and documents this conclusion and details the alternatives with discussions on: 1) background of the Cotton Valley play; 2) reservoir description; 3) analysis of present stimulation treatments and results; and 4) horizontal drilling as a completion alternative.

The following conclusions resulted from the study overviewed in this presentation:

* Commercial Cotton Valley Reef wells completed to date produce from a heterogeneous and diverse group of productive limestone reservoirs, with poor correlation between porosity and permeability. Thus, estimating individual well productivity from analysis of conventional open hole logs is difficult.

* Sixty-five percent of Cotton Valley fracture stimulations were determined to be unsuccessful.

* An accurate reservoir description that includes permeability and reservoir size is essential for optimum economic well stimulation. Chance of economic success could be increased to 80% from 35% by not stimulating wells with adequate permeability for their size, or those that require unrealistically long fracture half-lengths.

* Created fracture half-lengths were 32% to 48% less than designed, likely due to generation of complex fracture geometries - both near-wellbore and farfield - and/or poor proppant pack cleanup.

* For a base-case Cotton Valley well, best stimulation candidates are reservoirs with average permeabilities less than 0.15 md and estimated ultimate recoveries greater than 5.0 Bcf.

* Depletion of reservoir pressure prior to stimulation and application of horizontal drilling are possible completion alternatives.

BACKGROUND, PROBLEM DEFINITION

Over the last several years, there has been an active exploration effort focused on the discovery of gas reserves from the Upper Jurassic Cotton Valley Lime (Haynesville) interval in the East Texas basin.[1,2,3] Primary exploration objectives along this carbonate platform are overpressured, sour gas Cotton Valley reefs. Frantic land, 3-D seismic and drilling activity were fueled by some' outstanding early results that drove expectations to unrealistically high levels, e.g., average per well reserves of 70+ Bcf, initial potentials [greater than]25 MMcfd and a chance of 65% success using 3-D seismic.[1,4,5,6]

Brief history of play. Exploration and development of tight gas reserves from grainstone shoal facies of the Cotton Valley Lime interval have been ongoing within the East Texas basin since the 1950s. To date, about 632 Bcf has been produced from these grainstone facies. Fig. 1 highlights the Cotton Valley Lime production.

During the late 1970s and early 1980s, several companies attempted to extend grainstone production downdip to what was believed to be the shelf-ramp margin. In 1980, Texas Oil & Gas (TXO) discovered the first productive Cotton Valley Lime reef by drilling the McSwane 1 in Freestone County, an estimated 4 mi from known grainstone production. In 1982, a second reef was discovered with the TXO Marshall A-1 in Leon County, and the play was begun in earnest.

These two wells produced at unstimulated rates and pressures never before seen in the area. Early production rates averaged 7 MMcfd for the McSwane and 45 MMcfd for the Marshall, with estimated ultimate recoveries of 21 Bcf and 67 Bcf, respectively. Unfortunately, numerous dry holes were subsequently drilled, using 2-D seismic interpretation. Technology at the time proved to be inadequate to properly image small 40 to 100-acre carbonate buildups at depths greater than 15,000 ft, and the play stalled until 1993.

In 1993, Marathon, which had purchased TXO, drilled 218 ft of productive reef in the Poth 1, using advanced 3-D seismic interpretation. The well had an initial production rate of 20 MMcfd with an estimated ultimate recovery of 72 Bcf, and was responsible for igniting the current play.

By year-end 1998, about 3-million acres will have been leased, nearly 2,000 sq mi of 3-D will have been shot, and 90 wells will have been drilled at an estimated total cost of $1.1 billion. An estimated 533 Bcf of reserves have been discovered, to date, from 41 commercial wells within the proven trend. These wells were drilled to average depths of 15,400 ft, at dry hole costs of $5 million and completed costs of $8 million. The wells have about 180 ft of pay, original reservoir pressure of 13,200 psi and average maximum monthly flowrates of 15.6 MMcfd. The 41 commercial wells represent a chance of success of 51%, outstanding for an onshore exploration effort.

Stimulation problem development. As the play matured and industry's knowledge increased, expectations have fallen, but overall trend statistics remain attractive. Recently, to improve play economics, large hydraulic fracture treatments were routinely incorporated in the well completions. However, these have been largely unsuccessful for several reasons, including high treatment costs, high frac gradients, lack of a primary horizontal stress direction, highly heterogeneous reservoirs, poor barriers to frac height growth, and poor initial estimates of key reservoir parameters.

Fig. 2 indicates that most drilling/stimulation activity has occurred over the last several years, with all of the 36 frac treatments performed after 1994. Fracture stimulations have been incorporated into the completions to accelerate production and/or add reserves by "fracing into" previously isolated compartments. However, treatment analysis determined that only 35% of the stimulations were economically successful.

There are several reasons for the poor stimulation results. The primary problem was determined to be poor prefrac reservoir characterization. Chance of economic success could be increased to 80% from 35% by eliminating stimulation off 1) wells with adequate permeability for their size, and 2) wells that require unrealistically long fracture half-lengths.

The second major problem was that effective fracture half-lengths determined from pressure build-up analysis and production history matching were about one-third less than designed. Short, effective fracture half-lengths are probably due to the lack of strong barriers to vertical fracture height growth and the generation of a complex fracture geometry, both near-wellbore and farfield.

Of these underlying issues, only the problem of inadequate barriers to fracture height growth can realistically be attacked. For every 1.0 psi reduction in reservoir pressure, effective horizontal stress within the reservoir is decreased by about 0.66 psi. Therefore, depletion of reservoir pressure by just 1,500 psi prior to stimulation could result in a large enough stress contrast to significantly reduce fracture height growth.

As mentioned by Joshi,[7] "wherever large fracture extensions are difficult to achieve, long, horizontal wells provide an alternative completion option." Application of horizontal technology may be attractive for many Cotton Valley Reef completions. This has been supported by worldwide horizontal and specific field analogs, as well as analytical steady-state analysis, as will be discussed further.

GEOLOGIC OVERVIEW

The East Texas basin is one of four inland salt-diapir provinces of the northern Gulf Coast region. The four basins are thought to have begun as failed riffs associated with the opening of the Gulf of Mexico during the Early-Mid Mesozoic.[5] As Fig. 1 indicates, the East Texas basin is bound on the west and north by the Mexia Talco fault system, on the east by the Sabine Uplift, and joins the greater Gulf Coast Margin to the south.[8] Early rift activity within the basin was significant enough to allow marine waters to invade, evaporate, and leave southward-thickening salt wedges on an erosional surface developed over a Paleozoic basement.[5]

Subsequent movement of the Louann Salt, thermal cooling and its resulting subsidence, along with a rising sea level all combined to form favorable conditions for development of large carbonate buildups during the Late Jurassic.[8] Solving these complexities and placing them in a sequence stratigraphic framework to explain and predict elements of facies distribution will be key for future development.

The Cotton Valley Lime represents the downdip marine equivalent to the largely clastic Haynesville formation of Northern Louisiana.[5] The Lime comprises lime mudstones and wackestones, peloidal packstones, and oolitic and skeletal grainstones. The Smackover and Buckner formations underlie the Lime, while transgressive dark marine shales of the Bossier formation overlie it.[5] Generally, carbonate buildups completed to date have been restricted to the uppermost portion of the Cotton Valley Lime, with their tops completely encased by Bossier shales, which can be more than 1,500 ft thick.

Recent interpretation of 3-D seismic and subsurface rock data suggests a wide range of settings for development of productive carbonate buildups.[9] Fig. 3 is a block diagram of the East Texas basin which conceptualizes the perceived complex carbonate system of the Upper Jurassic in which grainstones, altered packstones, fractures and diverse reef assemblages contribute to the productive facies.[8] Use of 3-D has proven to be a necessary tool to accurately define and site Cotton Valley carbonate buildups.[8]

Additional information concerning the use of 3-D seismic and geological modeling in the exploration of Cotton Valley carbonate buildups is listed in the references.[1-11]

RESERVOIR CHARACTERIZATION

A proper reservoir description is essential for optimum economic completion of a Cotton Valley Reef well. The 80+ wells drilled to date in the Trend have encountered a wide variety of facies that have undergone different degrees and types of diagenesis. Formation of the carbonate buildups themselves appears to be cyclical and affected by small sea level changes.

The result is a heterogeneous and diverse group of productive limestone reservoirs that have poor correlation between porosity and permeability. Open hole logs have proven to be inadequate for estimating well productivity. Production testing, build-up analysis, and 3-D seismic interpretation are required for optimizing completion design. The following discussions describe relevant reservoir properties.

Deliverability. Average initial potential of the 41 commercial wells with available data is 14.2 MMcfd, at an average flowing tubing pressure of 4,900 psi. Average absolute open hole flowrate is 27.6 MMcfd, with an average maximum monthly flowrate of 15.6 MMcfd.

Reserves. Estimated ultimate recoveries of the 41 wells with available data were determined using decline curve analysis, P/Z analysis, and production history matching. All three methods gave similar results, with the average well's estimated ultimate recovery determined to be 13.0 Bcf. Fig. 4 is a graph of the lognormal distribution of estimated ultimate recoveries for the Trend based on decline curve analysis.

Areal extent. Cotton Valley carbonate buildups are small in areal extent, averaging just 66 acres, based on production history matching. This compares well with the average areal extent of 52 acres determined from 3-D seismic interpretation. Because a net pay estimate is required to production history match on drainage area, the good agreement suggests that net pay estimates from open hole logs are fairly accurate. Fig. 5 plots lognormal distribution of areal extent within the Trend based on production history matching. The large variation in areal extent results in a poor correlation between initial potential and estimated ultimate recovery.

Porosity and permeability. Fig. 6 crossplots porosity and permeability corrected to initial in-situ conditions. The data is from eight wells with 284 ft of combined full core and rotary sidewall core. As the figure indicates, there is poor correlation between porosity and permeability, e.g., a 6% porosity rock has a range of permeability from 0.004 md to 1 md, nearly a three-log-cycle spread. Therefore, open hole logs are inadequate tools for estimating individual well productivity, but are useful in calculating reservoir size.

The weak correlation between porosity and permeability is due to presence or lack of interconnected micro-porosity, vugular porosity, and natural fractures. The mean permeability based on a lognormal fit of the core data is 0.59 md, with an average porosity of 8%. Mean permeability based on 14 pressure buildups is 0.15 md, and the mean based on 31 production history matches is 0.21 md.

Pay. The average gross height of Cotton Valley reefs in the current trend is 480 ft. Average net pay using a 2% density porosity cut-off is 180 ft, which corresponds to a net-to-gross ratio of 38%.

Reservoir pressure. The Trend produces from highly overpressured reservoirs. Fig. 7 is a graph of reservoir pressure vs. depth that indicates abnormal pressures build at about 1.94 psi/ft, starting at 10,700 ft. Active generation of hydrocarbons and disequilibrium compaction are believed to be major causes of the abnormal pressures.[12,13]

Heterogeneity. Cotton Valley Lime buildups drilled to date exhibit heterogeneities at every level, from pore structures themselves to their overall distribution within the carbonate shelf. Heterogeneities are evident in SEM analysis, thin sections, formation imaging logs, core permeability-porosity crossplots, and geologic cross sections. An outstanding example of the complexity of these limestone reservoirs is illustrated in Fig. 8, in which open hole logs of the Bearcat 1, along with its two sidetracks, are shown. Intervals of greater than 6% porosity are highlighted. The porosity and gamma ray responses vary dramatically between all three wellbores and have no meaningful correlation, even though the sidetracks are only 65 ft and 365 ft apart.

STIMULATION RESULTS

Thirty-six of the 41 commercial wells have been stimulated. Of these, 23 had sufficient data to analyze the results, as summarized below.

Assumptions. Stimulation costs were estimated at roughly $1.0/lb of proppant, plus $100,000 for flow back and miscellaneous. A 15% discount factor was used for the net present value determination. Gas price was assumed to be $2.15/Mcf until 2001, $2.30 from 2001 through 2004, and then escalate at 3% after 2004. Stimulations were assumed only to accelerate production with no incremental reserve added. This final assumption was supported by several wells with sufficient prefrac/postfrac production to reliably compare prefrac/postfrac reserve estimates.

Typical stimulation. There have been large variations in stimulation designs, but data for an average stimulation is given in Table 1. In general, stimulations were designed to obtain fracture half-lengths of 400 to 600 ft and dimensionless fracture conductivities greater than 20, at an average cost of $500,000.

Table 1. Typical Cotton Valley Reef stimulation design

Design length:           400-600 ft
Gross height:            350 ft
Gel:                     60 lb HPG
X-linker:                Zirconate
Sand type:               20/40 Bauxite
Fluid volume:            250,000 gal
Sand volume:             400,000 lb
Sand ramp:               1-8 ppg
Pad size:                36%
Injection rate:          50 bpm
Average BHTP:            14,000 psi
Treatment down:          4-in.csg.
Average cost:            $500,000

Analysis. Estimated ultimate recoveries were primarily determined from decline curve analysis. Reservoir permeability was calculated from prefrac pressure buildup analysis or prefrac production history matching. When no prefrac production was measured, the rate was estimated using permeability calculated from postfrac pressure buildup analysis or postfrac production history matching. Created fracture half-lengths and conductivities were calculated from postfrac pressure buildup tests or postfrac production history matches. The postfrac net present value was subtracted from the prefrac net present value to determine whether the stimulation was successful.

Results. Table 2 lists results from analysis of the 23 stimulation treatments; only 35% were determined to be economically successful. Total cost for all 23 stimulations was estimated to be $11.4 million, with a total project net present value of $284,000. Average stimulation cost was $500,000, with an average net present value per stimulation of just $12,000. The average created fracture half-length was 191 ft. Average prefrac flowrate was 5.0 MMcfd; and average postfrac flowrate was 11.3 MMcfd, resulting in a productivity improvement factor (PIF) of 2.3.

[TABULAR DATA FOR TABLE 2 OMITTED]

DISCUSSION OF STIMULATION RESULTS

The overall chance of economic success for Cotton Valley Reef stimulations is 35%. Two major reasons for this poor performance are: 1) lack of an accurate reservoir description prior to stimulation, and 2) generation of effective fracture half-lengths that are one-third shorter than designed.

Effect of poor reservoir characterization. Recently, most completions have been designed to fracture stimulate wells prior to any testing. The stimulation is incorporated into the initial completion to reduce flowback costs, reduce downtime and add reserves by fracing into isolated, compartmentalized reservoir portions. However, knowledge of reservoir permeability and ultimate recovery is essential to prefrac design and economic success.

Table 3 lists reservoir parameters for a base-case well. Figs. 9 and 10 show the effect that permeability and reservoir size have on the half-length required to achieve economic success, and the probability of achieving that half-length. As Fig. 9 indicates, for a base-case reef well, the best stimulation candidates are those with less than 0.15-md permeability. When reservoir permeability exceeds 0.15 md, a 23-acre reef is already depleting rapidly. The PIF required to achieve economic success increases with increasing reservoir permeability, while the probability of achieving the required productivity improvement decreases.

Table 3. Base-case Cotton Valley Reef well

Median res. perm.:              0.085 md
Median areal extent:            23 acres
Net pay:                        180 ft
Average porosity:               8%
Permeability anisotropy:        7
Perforation midpoint:           14,979 ft
Avg. initial res. press.:       13,200 psi
Water saturation:               10%
Skin:                           0
Tubing size:                    3.5 in.
Wellbore radius:                0.21 in.
Avg. res. temp.:                340 [degrees] F
Total compressibility:          9.6 x [10.sup.-5]
Median EUR:                     4.2 Bcf

A similar effect is evident for reservoir size, as illustrated in Fig. 10. A base-case reef well with an estimated ultimate recovery of less than 5 Bcf requires a large productivity improvement factor and has a low chance of achieving the required fracture half-length. A small reservoir with a permeability of 0.085 md already depletes very rapidly. Significant production acceleration is required to make the stimulation of a small Cotton Valley carbonate buildup economic.

Overall results were substantially improved by eliminating the stimulation of wells with adequate permeability for their size and wells that required greater than 300 ft of effective fracture half-length. Only 10 of 23 wells actually required stimulation when using this criteria. The chance of economic success for this subset of wells improves to 80% from 35%, and total stimulation cost drops to $3.8 million from $11.4 million. Total net present value improves 15 fold, increasing to $4.3 million from $284,000. The average productivity improvement factor increases to 2.7 from 2.3.

Therefore, knowledge of reservoir permeability and reservoir size prior to stimulation is essential. PIF, corresponding fracture half-length, and probability of achieving required fracture half-length are all based on estimates of reservoir permeability/size. Currently, this data cannot be adequately estimated from open hole logs and 3-D seismic interpretation. Both a prefrac production test and a pressure buildup test are recommended prior to design/implementation of any stimulation procedure.

Problems with creating effective fractures. In general, stimulations have been designed to achieve effective fracture half-lengths of 400 to 600 ft. Actual average effective fracture half-length is 191 ft, or 32% to 48% less than designed. Fig. 11 illustrates the probability of achieving or exceeding a desired half-length. The probability of achieving a 400-ft half-length is less than 15% and less than 5% for a 600-ft half-length. The probability of achieving any kind of effective fracture half-length is only 75%.

Shorter-than-designed half-lengths are likely due to generation of a complex fracture geometry, both near-wellbore and farfield. High frac gradients of near 1.0 psi/ft, no clear primary horizontal stress direction, poor barriers to fracture height growth, and a highly heterogeneous reservoir may all contribute to a complex created-fracture geometry. Any of these characteristics alone are significant issues; combination only increases the likelihood of a problem. Inability to adequately pressure history match many of the frac stimulations is evidence of this complexity.

Possible solutions to help solve some of the near-wellbore issues include increasing injection rate, increasing viscosity, pumping heavy gel or sand slugs during the pad, decreasing pad volume or decreasing overall job size. Farfield issues of reservoir heterogeneity and lack of a primary horizontal stress direction cannot realistically be altered. However, the frac gradient can be reduced, and fracture height containment can be improved by producing the reservoir prior to stimulation.

For every 1.0-psi reduction in reservoir pressure, stress contrast between overlying and underlying beds can be increased by about 0.66 psi.[14] Using the rule of thumb of at least 1,000 psi of stress contrast between beds, a 1,500-psi minimum reduction in average reservoir pressure is required. Additional benefits include achieving a good prefrac production test and reducing required stimulation horsepower.

Generation of poor effective half-lengths may also be due to inadequate proppant pack cleanup. Recent experiments at STIMLAB indicate that fracture cleanup may not be as efficient as once thought.[15] Inertia forces required to clean out gel residue in the proppant pack are probably more than sufficient for most Cotton Valley Reef wells. However, the wells produce nearly liquid-free and therefore do not have any water in addition to that which is pumped during stimulation to help clean up gel residue. Lack of producible water may reduce the ability of fractures to clean up.

HORIZONTAL DRILLING AS AN ALTERNATIVE

Horizontal wells can outperform vertical completions in areas in which large fracture extensions are difficult to achieve? Worldwide and specific field analogs and an analytical steady-state analysis were used to evaluate the application of horizontal drilling. In many cases, a horizontal well may outperform a vertical fractured well. Estimated incremental cost above a vertical fractured well is $729,000 to achieve a 2,000-ft lateral.

Worldwide analogs. based on performance of more than 1,300 horizontal wells in different reservoirs throughout the world, the average horizontal well produces five times the rate of an unstimulated vertical well.[16] PIF probability distributions for several reservoir types and for fractured vertical reef wells are shown in Fig. 12 (modified from Flumerfelt).[16] At a 60% chance of occurrence, the horizontal PIF is one full unit higher than for a fractured vertical well, and increases to a 4-unit improvement at a 20% chance of occurrence.

Specific field analogs. Rospo Mare field in the Adriatic Sea and the Niagaran Reef Trend of Northern Michigan are good field analogs.[17,18,19] Rospo Mare, operated by Elf Aquitaine Production Co., is a karsted limestone reservoir that was developed with horizontal wells. The reservoir is highly heterogeneous, with natural fractures evident in cores and outcrops. 18 Elf has published results indicating PIFs greater than 20, with reserve increases of 59%.[18]

The Niagaran Reef Trend has recently been revitalized by use of short-radius horizontal drilling. 17 Reservoir characteristics are very similar to those of the Cotton Valley. Areal extents range from 30 to 1,000 acres, with an average of 100 acres. Gross reservoir height is 250 to 600 ft, with an average 150-ft net pay. Shell presented results that indicated PIFs of two to three from short-radius horizontal wells with average lateral lengths of 600 ft.[17] The Gas Research Institute presented results from a field experiment that reported a PIF of five, and a reserve increase of 6%, for a 2,000-ft horizontal Niagaran Reef well.[19]

Analytical solution. The productivity improvement factor for horizontal wells is highly dependent on lateral length, net pay thickness and horizontal-to-vertical permeability ratio.[7] Sensitivities to these parameters were run on a base-case Cotton Valley Reef well shown in Table 3. The base-case horizontal-to-vertical permeability anisotropy of seven was determined from a build-up analysis on a partially penetrated reef well. Sensitivities are shown graphically in Fig. 13.

Base-case PIF was determined to be 2.7, which is 17% higher than the average fractured vertical reef well. Horizontal PIF is higher than that of a fractured vertical reef well when: net pay is 200 ft or less, lateral length is 2,000 ft or greater, and horizontal-to-vertical permeability anisotropy is 40 or less. The horizontal PIF is double that of an average fractured vertical reef well when net pay drops below 100 ft, or horizontal-to-vertical permeability anisotropy is less than two.

Biggest risks with horizontal wells. A good estimation of horizontal-to-vertical permeability anisotropy is essential to predict well performance, but this is extremely difficult to determine prior to drilling. The internal structure of the Cotton Valley Reefs, based on the geologic growth model of stacked reef cycles, could support the concept of poor vertical permeability. A permeability anisotropy greater than 40 is uneconomic in almost any probable Cotton Valley Reef reservoir. Additional problems are associated with the drilling of horizontal wells at the depths, pressures, and temperatures of the existing play.

ACKNOWLEDGMENT

The author thanks John Abeln of Broughton Resources; Raymond Flumerfelt of Holditch & Associates; Nathan Meehan, Dave Kyte, Steve Blanke, Gary White, Ray Walker, Casey Clawson, and Janet Hill of Union Pacific Resources and Melissa Richardson, for help in reviewing and editing the paper; and Olga Melendez of Union Pacific Resources for her help on many of the figures. This article is adapted from paper SPE 49049, "Completion and stimulation practices in the prolific Cotton Valley Reef Trend of East Texas," written by the author and presented at the 1998 SPE Annual Technical Conference and Exhibition, New Orleans, Louisiana, Sept. 27-30, 1998.

LITERATURE CITED

1 Wheatley, R., "East Texas Cotton Valley Pinnacle Reef E&D emerging as hot gas play," Oil & Gas Journal, March 1997, pp. 23-26.

2 Montgomery, S., et al., "East Texas Upper Jurassic Reefs and an expanding Cotton Valley Lime play," Oil & Gas Journal, September 1997, pp. 82-86.

3 Hatch, K. D. "The Cotton Valley Reef Trend, Robertson, Leon and Freestone Counties, Texas," Abstract from Houston Geological Society Meeting, January 1996.

4 Williams, P. "Cotton Valley Reefs," Hart's Oil & Gas News, April 1997.

5 Williams, P. "Getting better all the time," Hart's Oil & Gas News, June 1997.

6 Montgomery, S., Petroleum frontiers. Vol. 10, No. 2, Petroleum Information Corp., Littleton, Colorado, 1993.

7 Joshi, S. D., Horizontal well Technology, PennWell Publishing Co., Tulsa, Oklahoma, 1991, p. 535.

8 White, G., et al., "Salt tectonism in the East Texas basin. Implications for Cotton Valley Reef distribution," 1998 AAPG Annual Meeting abstract.

9 Tarkington, D. K., et al., "Characteristics of a new productive Cotton Valley buildup trend, East Texas basin," 1998 AAPG Annual Meeting abstract.

10 Montgomery, S., Petroleum frontiers. Vol. 13, No. 4, Petroleum Information/Dwights LLC, Denver, Colorado, 1997.

11 Goldhammer, R. K., "Second order accommodation cycles and points of stratigraphic turnaround: Implications for high-resolution sequence stratigraphy and facies architecture of the Haynesville and Cotton Valley Lime Pinnacle Reefs of the East Texas Salt basin," WIGS Bulletin, Vol. 37, No. 7, March 1998.

12 Law, B. E., et al., "Geology of right gas reservoirs in the Pinedale Anticline area, Wyoming, and at the Multiwell Experiment Site," U.S. Geological Survey Bulletin, 1990.

13 Osborne, M. J., et al., "Mechanisms for generating overpressure in sedimentary basins: A re-evaluation," AAPG Bulletin, June 1997, pp. 1023-1041.

14 Economides, M. J., et al., Reservoir stimulation, Second Edition, Prentice Hall, Englewood Cliffs, New Jersey, 1989.

15 Anon., STIMLAB Proppant Consortium, 1997.

16 Flumerfelt, R. W., Jr.," Screening criteria for horizontal well candidate selection," Petroleum Engineer International, Hart Publications, November 1997, pp. 55-58.

17 Lanier, G. H., "Low-cost short radius re-entry horizontal drilling program revitalizes aging Northern Michigan Niagaran oil fields," paper SPE 36482, presented at the 1996 SPE Annual Technical Conference and Exhibition, Denver, Colorado, Oct. 6-9, 1996.

18 Gauchel, R., et al., "Rospo Mare field: A unique experience of heavy oil production with horizontal wells in a karst reservoir in presence of a strong tilted hydrodynamism," paper SPE 36869, presented at the 1996 SPE Annual Technical Conference and Exhibition, Denver, Colorado, Get. 6-9, 1996.

19 McDonald, J. W., et al., Howell storage field horizontal well field experiment, Gas Research Institute Topical Report, May 1995.

The author

Michael F. Richardson, senior engineer, Union Pacific Resources, Ft. Worth, Texas, graduated from Texas A&M University with a BS in petroleum engineering in 1989. He worked as a completions/stimulation engineer for Mobil E&P in West Texas. He has spent several years supporting the Cotton Valley Reef effort as an exploration engineer, plus two years on the tight gas sands of the East Texas bas/n as a development/operations engineer, and three years on the Austin Chalk Business Unit and one year as an operations engineer in the South Texas Business Unit. Mr. Richardson is a member of SPE and AADE.

COPYRIGHT 1998 Gulf Publishing Co.
COPYRIGHT 2000 Gale Group