Basin-Center Gas Potential In Bossier Shale

A basin-center, continuous-gas accumulation might have been discovered recently in sandstones within the Bossier Shale, the lower formation of the Cotton Valley Group. In a currently developing play on the western flank of East Texas Basin, gas is being produced from turbidite sandstones within the Bossier Shale. These turbidite sandstones probably are downdip time-equivalent deposits of deltaic sandstones in the lower portion of the Cotton Valley Sandstone and reportedly were deposited seaward of the underlying Haynesville carbonate platform edge in a slope or lowstand-fan setting. Accommodation space was provided by salt withdrawal such that updip and lateral traps currently are formed by pinchout of sandstone into shale. Two stacked, stratigraphically separate Bossier turbidite-fan systems occur at depths of 13,000 to 14,000 feet. Two fields, Dew and Mimms Creek, with combined estimated recoverable reserves of more than one TCFG, currently are being developed by Anadarko Petroleum, one of the main operators. As of January 2000 (PI-Dwights Drilling Wire, Jan 3, 2000; Jan 12, 2000), Anadarko had drilled more than 100 wells with only one dry hole in this Bossier sandstone play. Gas-charged sandstones reportedly are overpressured, and no water has been encountered in the system (Exploration Business Journal, 2nd quarter, 2000). Within the upper turbidite-fan interval, porosity ranges from 6 to 15 percent and permeability from 0.01 to 1.0 mD. Initial production rates from wells average 3 to 4 MMCFGD after fracture stimulation and decline exponentially with estimated per-well recoveries of 1 to 5 BCFG. In the lower sandstone interval, porosity ranges from 9 to 20 percent, permeability from 1 to 10 md, pressures are higher, and initial production rates of up to 30 MMCFGD have been obtained. This play does not seem to involve the classic type of basin-center gas accumulation with trap produced by inherent low-permeability of reservoir sandstones. Instead, the trap seems to be provided by marine shales that completely encase these turbidite sandstones, but the sandstone reservoirs are overpressured, seem to lack water and gas-water contacts, and are gas-charged over an extensive area as witness by only one dry hole in more than 100 wells drilled.

Geological/Petrophysical Overview

The Bossier sands are part of the Upper Jurassic-age Cotton Valley Group deposited in the East Texas Basin. The Bossier interval, which lies immediately beneath the Cotton Valley Sandstones, is a thick, lithologically complex system containing black to gray-black shales interbedded with fine-grained to very fine-grained argillaceous sandstones. The Cotton Valley Group is underlain regionally by the Upper Jurassic Louark Group, which includes other hydrocarbon-bearing formation such as the Smackover Carbonates and Haynesville/Cotton Valley Limestones. Overlying the Cotton Valley Group is the regionally productive Lower Cretaceous Travis Peak and Pettit formations.

Bossier sand rock types include clean sandstones, argillaceous weakly laminated sandstones, dolomite sandstones, and argillaceous burrowed siltstones. Intergranular constituents are primarily quartz overgrowths, diagenetic clays in the sands, and detrital clays found in both sand and silt. The clay fraction is predominantly grain coating chlorite and illite. Bossier sands also have a narrow range of grain size, typically from upper very fine to fine. The sands are medium to well sorted, while the silts are poorly sorted. Bossier sands also exhibit a significant diagenetic overprint, including mechanical compaction, cementation from quartz overgrowths, grain-coating/pore-lining clay development, and grain dissolution. Figure 2 contains a type-log containing the four target sands for the project: York, Bonner, Moore, and Bossier Marker.

Although effective porosity in the Bossier sands varies from 1% to 17%, the average porosity in the net sand ranges from 6% to 10%. Absolute permeability varies from 0.001 md to 1 md. Average permeability in the reservoir rock averages 0.0005 md to 0.05 md. Measured water saturation in the reservoir rock ranges from as low as 5% in the most permeable rock to as high as 50% in the lower-quality reservoir rock.

From: Advanced Fracturing Technology For Tight Gas: An East Texas Field DemonstratioN. Final Report October 1, 2001 December 31, 2004, March 2005. http://www.osti.gov/bridge/servlets/purl/861428-O85aKa/

Bossier Background

The current Bossier play is located on the western flank of the East Texas Basin. Figures 1 provides a cross section for the Dowdy Ranch Field, where the wells involved in this study are located. Although gas has been produced from the Bossier interval from the 1970’s, the current play began in 1996 and gained major attention with Anadarko’s

aggressive 1998 drilling program. Anadarko has drilled extensively in this area and no significant drilling problems have been encountered. Anadarko is the most active operator in the Bossier play with operators such as Cross Timbers and Pioneer also being very active.

Bossier wells generally produce dry gas with little or no water production from sands embedded in the Bossier. The water saturation in some sands has been measured at below 10%. This is consistent with a “basin bottom” gas column more than 3000 feet thick reported by Montgomery1. Most production is from over-pressured zones that are

part of a regional overpressure cell that extends over the southern half of the East Texasbasin (Montgomery and Karlewicz2). The Bossier sands are part of the upper Jurassic age Cotton valley group deposited in the East Texas basin. The Bossier interval, which lies immediately beneath the Cotton Valley sandstones, is a thick lithologically complex

system containing black to gray-black shales interbedded with fine-grained argillaceous sandstones. The Cotton Valley group is underlain regionally by the Upper Jurassic Louark group, which includes other hydrocarbon bearing formations such as the Smackover carbonates and Haynesville/Cotton Valley limestones. Overlying the Cotton Valley group is the regionally productive lower cretaceous Travis peak and Petit formations.

Productive sands are found at depths ranging from 12,000 to 15,000 feet. A stratigraphic column (Montgomery and Karlewicz2) is shown in Figure 2. The Bossier is the time-equivalent of the fluvial-deltaic Cotton Valley sands that produce to the west. The Bossier sands were deposited in paleo-bathymetric lows formed by salt movement contemporaneous with Bossier deposition. Deposition was also influenced by carbonate buildups and shoals in the underlying Cotton Valley lime. The sand transport feeding the system came both from southwest and east-southeast trending channels. The Bossier sands are generally classified as low stand delta front and pro delta material. The sands occur in the top 500-600 feet of the Bossier shale. They are found as lenses, segregated pods, and channel trends. The areal extent of these sands range from hundreds of acres to several square miles, with thicknesses up to hundreds of feet. As drilling has continued on tighter spacing the heterogeneity of the sands has become more apparent. Five separate sand units (Taylor, Shelly, Moore, Bonner, and York) have been identified (see Figure 2). The hydrocarbons in the Bossier sands are believed to be sourced from the surrounding shale.

The upper three sands, Taylor, Shelly, and Moore, are generally of lower reservoir quality than the Bonner and York. They have reported porosities in the 6 to 15% range and permeabilities in the 0.001 to 1 md range. They generally have higher clay content than the Bonner. The porosity in the Bonner sand varies from 8 to 20%, with permeabilities in the range of sub-millidarcy to several millidarcies (normally less than 0.1 md). The sands are generally quartz arenites or subarkoses. They are moderately sorted with subangular to subrounded grains. In the productive zones the majority of the storage seems to be in lower permeability sands, which in part probably accounts for the observed hyperbolic decline curves. We primarily focused on the Bonner and York sands in this study, as those are the predominant pay zones in the Dowdy Ranch Area. Bossier sand rock types include clean sandstones, argillaceous weakly laminated sandstones, dolomitic sandstones and argillaceous burrowed siltstones. Intergranular constituents are primary quartz overgrowths, diagenetic clays in the sands, and detrital clays found in both sand and silt. The clay fraction is predominantly grain coating chlorite and illite. Bossier sands also have a narrow range of grain size, typically from upper very fine to fine. The sands are medium to well sorted, while the silts are poorly sorted. Bossier sands also exhibit a significant diagenetic overprint, including mechanical compaction, cementation from quartz overgrowths, grain coating/pore lining clay development and grain dissolution.

Although the effective porosity in Bossier sands varies from 1% to 17%, the average porosity in the net sand ranges from 6% to 10%. Absolute permeability varies from 0.001md to 1 md in all the rock types, while average permeability in the reservoir rock ranges from 0.005 md to 0.05 md. Measured water saturation in the reservoir rock ranges from as low as 5% in most cases to as high as 50% in the lower-quality reservoir rock. All producing wells drilled in the Bossier play are hydraulically fractured. The development of optimal fracturing procedures, therefore, has a big impact on the long term economic viability of the play.

Introduction

The primary objective of this research was to improve completion and fracturing practices in gas reservoirs in marginal plays in the continental United States. The Bossier Play in East Texas, a very active tight gas play, was chosen as the site to develop and test the new strategies for completion and fracturing. Figure 1 provides a general location map for the Dowdy Ranch Field, where the wells involved in this study are located. The Bossier and other tight gas formations in the continental Unites States are marginal plays in that they become uneconomical at gas prices below $2.00 MCF. It was, therefore, imperative that completion and fracturing practices be optimized so that these gas

wells remain economically attractive. The economic viability of this play is strongly dependent on the cost and effectiveness of the hydraulic fracturing used in its well completions.

Water-fracs consisting of proppant pumped with un-gelled fluid is the type of stimulation used in many low permeability reservoirs in East Texas and throughout the United States. The use of low viscosity Newtonian fluids allows

the creation of long narrow fractures in the reservoir, without the excessive height growth that is often seen with cross-linked fluids. These low viscosity fluids have poor proppant transport properties. Pressure transient testsrun on several wells that have been water-fractured indicate a long effective fracture length with very low fracture conductivity even when large amounts of proppant are placed in the formation. A modification to the water-frac stimulation design was needed to transport proppant farther out into the fracture. This requires suspending the proppant until the fracture

closes without generating excessive fracture height. A review of fracture diagnostic data collected from various wells in different areas (for conventional gel and water-fracs) suggests that effective propped lengths for the fracture treatments are sometimes significantly shorter than those predicted by fracture models.

There was no accepted optimal method for conducting hydraulic fracturing in the Bossier. Each operator used a different approach. Anadarko, the most active operator in the play, had tested at least four different kinds of fracture treatments. The ability to arrive at an optimal fracturing program was constrained by the lack of adequate fracture models to simulate the fracturing treatment, and an inability to completely understand the results obtained in previous fracturing programs. This research aimed at a combined theoretical, experimental and field-testing program to improve fracturing practices in the Bossier and other tight gas plays.

Executive Summary

This DOE sponsored project at the University of Texas at Austin and Anadarko Petroleum Corporation conducted over the past 2 years has focused on strategies for improving productivity in the Bossier play in East Texas. The key to the successful application of successful fracture designs is the selection of a specific strategy and fracture design for a specific well based on an analysis of the well data. The development of 3-D hydraulic fracture models that are capable of accurately modeling proppant transport and fracture propagation is crucial to the successful selection of these strategies in tight gas formations over a broad cross-section of tight gas sand assets with widely varying stress regimes and formation properties.

In summary, laboratory and modeling work conducted at University of Texas as well as field experience in Anadarko wells in the Bossier have resulted in improvements to our understating of proppant placement and fractured well productivity in tight gas sands. By conducting hydraulic fracture simulations that incorporate realistic models for

proppant transport and fracture propagation, such as those developed here, the selection of an optimal fracturing strategy can be speeded up and performance predictions made with a greater degree of confidence. Without such detailed studies many of these strategies which may be very effective in some locations may prove to be completely ineffective in others.

Major Operator Update

Anadarko

Activities at the Company’s properties in east Texas are concentrated in the Bossier play with production and development activities in the Dowdy Ranch, Dew/Mimms Creek, Bald Prairie, Beargrass, Holly Branch and Marquez fields. In 2006 Anadarko Bossier play produced 183 MMcf/d natural gas n, 31 total MBOE/d, 1063 producing wells, 38 wells drilled in 2006 with 100% success rate.We've been operating in the greater Bossier trend since the mid-1990s. The region is distinguished by high-quality, unconventional gas wells. 287,000 net acres,, 2.4 TCF estimated ultimate recovery (gross).

Chesapeake

Deep Bossier - we have recently expanded our position and increased our exposure to the emerging Deep Bossier play in east Texas through our own leasing efforts and an investment in and partnership with Houston-based Gastar Exploration (AMEX: GST). We currently own approximately 350,000 net leasehold acres in the play and are drilling our first wells with acceleration of the program expected in 2007.

ConocoPhillips

In the Bossier Trend in east Texas, the company had interests in more than 177,000 net acres at year-end 2005 and is expanding beyond its Savell field development with other exploration and development activities along the trend. Net production averaged 76 MMCFD of gas, 7 MBOD and 1 MBD of NGL in 2005.

EnCana

Natural gas, east Texas, acquired in 2004. This resource play targets the Deep Bossier and Cotton Valley zones. In 2006, EnCana increased its working interest in the Deep Bossier play, increasing total land holdings to 170,000 net acres. During 2006, EnCana drilled approximately 59 net wells in the resource play. Natural gas production for the resource play averaged approximately 99 million cubic feet per day (MMcf/d).

Gastar

In April 2007, Gastar announced the sale of a portion of Gastar’s undeveloped deep Bossier leasehold in Robertson County, Texas for $92 million in cash proceeds including the acquisition by the buyer of 10 million newly issued Gastar common shares at a price of $2.00 per share. The transaction resulted in Gastar reducing its net acreage leasehold from approximately 26,000 acres to approximately 16,350 acres. Gastar has identified approximately 135 deep Bossier drilling locations containing an estimated 350 BCF of net reserve potential. In addition, Gastar has also identified another 100 BCF of gross reserve potential associated with shallower formations that could also be explored with the Bossier drilling program.

XTO Energy

XTO Energy is a prominent producer in East Texas and northern Louisiana, with the Freestone Trend leading the way. Current Freestone Trend daily production of about 585 MMcfe and production will continue to grow at a pace of 10 - 20% per annum over the next few years. XTO has identified up to 2,100 drilling locations in the Eastern Region, which represent 4 to 5 years of development spanning from East Texas into northern Louisiana. The primary producing horizons in the region include the Pettit, Rodessa, Travis Peak (Hosston in La.), Cotton Valley Sand, Bossier (Gray Sand in La.), Cotton Valley Lime and the Haynesville, ranging in depth from 5,000 to 13,000 feet.