EXPLORATION New basins invigorate U.S. gas shales play

Jan. 22, 1996
Scott R. Reeves Vello A. Kuuskraa Advanced Resources International Inc. Arlington, Va. David G. Hill Gas Research Institute Chicago While actually the first and oldest of unconventional gas plays, gas shales have lagged the other main unconventional gas resources-tight gas and coabed methane-in production and proved reserves. Recently, however, with active drilling of the Antrim shales in Michigan and promising results from the Barnett shales of North Texas, this gas play is growing in
Scott R. Reeves
Vello A. Kuuskraa

Advanced Resources International Inc.
Arlington, Va.
David G. Hill
Gas Research Institute
Chicago

While actually the first and oldest of unconventional gas plays, gas shales have lagged the other main unconventional gas resources-tight gas and coabed methane-in production and proved reserves. Recently, however, with active drilling of the Antrim shales in Michigan and promising results from the Barnett shales of North Texas, this gas play is growing in importance.

While once thought of as only an Appalachian basin Devonian-age Ohio shales play and the exclusive domain of regional independents, development of gas shales has expanded to new basins and has begun to attract larger E&P firms. Companies such as Amoco, Chevron, and Shell in the Michigan basin and Mitchell Energy & Development and Anadarko Petroleum Corp. in the Fort Worth basin are aggressively pursuing this gas resource.

With the success seen in these basins, E&P companies are also testing the potential of gas shale plays in the New Albany shales of the Illinois basin, the Niobrara shales of the DJ basin, and Woodford shales in southern Oklahoma the (Fig. 1.[ 57309 bytes])

Industry status

At yearend 1995, gas shales provide an estimated 800 MMcfd (300 bcf/year) of U.S. natural gas production. The principal producing areas are the traditional Devonian shales of the Appalachian basin as well as the emerging Antrim shales of the Michigan basin and the Barnett shales of the Fort Worth basin (Fig. 2[21621 bytes]). Overall, production from gas shales has more than doubled since 1990 and is up significantly since 1992, the year tax credits ended, the Antrim shales of Michigan fueling the bulk of the growth with (Table 1[13584 bytes]).

The other vital statistics for the "gas shale industry" are provided (Tables 2 [14973 bytes], 3 [13858 bytes] and 4 [13975 bytes ]). The data include information from the Appalachian basin, a region for which it has traditionally been difficult to obtain data. These statistics show that the gas shales industry is vigorous and has been able to continue to grow without tax subsidies.

  • Gas shale drilling has rebounded strongly in 1995 with 1,300 completions after a drop to about 900 completions in 1993 and 1994 (Table 2 [14973 bytes]). Active drilling in the Antrim shale-with an estimated 850 completions, nearly double the rate of the prior 2 years-has led the way. The continuing decline in Appalachian basin gas shale drilling has been counterbalanced by increased activity in the Fort Worth and Illinois gas shale basins.

  • Currently, about 22,000 gas shale wells are active and producing, with the bulk of these still in the Appalachian basin (Table 3 [13858 bytes]). The major recent growth has been in the Antrim shale, where an estimated 4,600 wells are producing at the end of 1995, up from 2,043 wells at yearend 1992. This growth has countered the decline in producing gas shale wells in the Appalachian basin as low gas prices have led operators to shut in nearly 2,000 marginal wells.

  • While the resource in place has been estimated in the hundreds of trillion cubic feet, currently only about 3.3 tcf of the gas shale resource is booked as proved reserves, principally in the Appalachian and Michigan basins. At the end of 1994, proved reserves of gas shales were up by 28% from 1992, when the tax credits expired.

  • After many years of declining returns to drilling, reserves per well have steadily improved for gas shales as operators have shifted more of their attention to the emerging gas shale basins. With the active recent drilling and the improvements noted in gas recovery per well in these basins, proved reserves of natural gas from shales should be up strongly in 1995.

Gas shale well productivity varies widely by basin, due to variations in the intensity of the local natural fracture network, richness of the shale, net thickness, and other key reservoir properties (Table 5 [13598 bytes]). In addition, changes in well completion and operating methods have led to increased per well reserves and productivity. Currently, the deep (7,000-8,000 ft) wells in the Barnett shale are the most prolific, showing initial production rates of 500-1,000 Mcfd. Next come the wells drilled into the "sweet spots" of the shallow (1,500 ft) Michigan basin that peak at 300-500 Mcfd after dewatering. At the other extreme are the moderate depth (4,000 ft) Devonian-age shale wells of the Appalachian basin that produce a few hundred thousand cubic feet per day initially and rapidly decline to below 100 Mcfd. Even here, large 1,000 Mcfd wells are occasionally found, particularly when the wells intersect extensive gas filled natural fracture systems.

Improved well completions using nondamaging methods and more effective hydraulic stimulations have raised the overall per well reserves for newly drilled wells. As important has been the improved ability (or good fortune) to find highly naturally fractured areas that offer the potential for several fold increases in gas production and per well reserves. Together, these two technologies have led to a 50% increase in gas shale reserves per well, as presented in Part 1 of this series (OGJ, Dec. 11, 1995, p. 76).

Table 6 [17753 bytes] shows the latest costs and economics for two of the emerging gas shale plays, the Michigan (Antrim) and the Fort Worth (Barnett) under "sweet spot" and average well conditions, further emphasizing the importance of exploring for naturally fractured settings.

Basin activity Appalachian basin

Much of the outlook on the overall gas shale resource has been colored by the long experiences with the Devonian-age shales of the Appalachian basin. Over 20,000 shale gas wells have been drilled in this basin, with about 16,000 of these still producing. Well drilling that averaged about 500 wells/year in the early 1990s has remained reasonably strong at 300-400 wells/year since the end of the tax credits. Historical development has been focused in western West Virginia, eastern Ohio, and northeastern Kentucky; however, most of the recent activity in the Appalachian basin has been centered around Pike County, Ky., in Big Sandy field. Three of the big operators in the Appalachian basin include Equitable Resources Exploration (EREX) with more than 3,000 producing shale wells, Ashland, also with approximately 3,000 producing wells, and Colombia Gas, with nearly 2,000 wells.

The Devonian-age shales of the Appalachian basin span from southwestern New York to eastern Kentucky and central Tennessee. The Devonian-age rocks are composed of a series of interbedded shales, siltstones, and sandstones. The entire Devonian sequence thins and becomes more of an organic-rich black shale to the south and west, although gas production from the Devonian has been targeted throughout the Appalachian basin. The reservoir varies from region to region, and gas can be found in the tight sand units, siltstones, fractured shale sections, and the organic rich black shales. Many operators commingle production from the black shale interval with siltstone and tight sand production from shallower horizons such as the Berea and Weir formations.

The major change that is affecting the outlook for this gas play is the decline in productivity of the more recently drilled wells. Based on data provided by Columbia Natural Resources to the National Petroleum Council, the "recent" wells (drilled since 1971) are averaging about 200 MMcf in terms of ultimate recovery, down from 680 MMcf of ultimate recovery for "older" (pre-1971) wells (Table 7 [13349 bytes]). The change in Appalachian gas shale well drilling and ultimate recovery, provided below, shows particularly significant de-clines for Kentucky and Ohio. In Kentucky, the per well reserves in the Big Sandy area (Pike, Martin, and Floyd counties) dropped to 0.24 bcf/well from 0.84 bcf/ well. The results from recent drilling in the Ohio portion of the Big Sandy (Lawrence County), were even more dismal, where new wells reserves were only one-tenth the previous 0.5 bcf/well. One explanation for the lower recoveries by the new wells is that the bulk of the new wells are infill wells drilled into areas of lower pressures partly in areas depleted by older wells.

It appears the major challenges facing producers in the basin are to: (1) identify new undrilled areas underlain by good gas content and intensive natural fractures; (2) rigorously establish the drainage and depletion of existing (old) wells for more productive infill development; and, (3) develop improved well completion and stimulation methods to link more of the vertical pay to the wellbore.

More recent data indicate that with the lower number of better completed and more selectively placed wells, productivity for the Applachian basin gas shales has rebounded to 0.3-0.5 bcf/well for 1990 through 1994.

Michigan basin

The Antrim shale has become one of the most active gas plays in the U.S. During 1992, Otsego County, Mich., led the nation in well drilling, and total Antrim shale completions represented nearly 15% of all U.S. gas well completions as operators rushed to drill prior to tax credit expiration. Following this surge, Antrim drilling levels in 1993-94 fell to about one half the 1992 peak. In 1995, however, Antrim activity is up sharply as operators are drilling in new areas of the basin and recompleting additional zones in existing wells.

The Antrim shale is an organic-rich, black shale deposited during the Devonian and Mississippian ages in an area of approximately 30,000 sq miles of Michigan and northern Indiana. The Antrim shale productive trend of upper Michigan is found at depths ranging from 1,009-2,000 ft. The Antrim is divided into four primary members-the undifferentiated Upper Antrim and the Norwood, Paxton, and Lachine members of the Lower Antrim. The Norwood and Lachine shale members are the main pay intervals. Early projects seeking to tap the free gas at the top of the pay targeted only the upper portion of the Lachine shale member. In the last several years, many wells have been deepened to the Norwood, and most new wells are now drilled through Norwood and dually completed in Norwood and Lachine.

Twenty-two operators are active in the Antrim gas play, mostly in the original productive trend in Otsego, Antrim, and Montmorency counties. The largest operators in the Michigan basin are Terra Energy, who in 1994 produced 27 bcf from 905 wells (Table 8 [12605 bytes]), followed by Ward Lake Drilling and Wolverine Environmental Production. A major change occurred in 1995 with the takeover of the two largest operators in the basin. Nomeco, a subsidiary of CMS Energy, purchased Terra Energy, and Belden and Blake Corp., now the largest operator in the eastasn U.S. in terms of number of wells, purchased Ward Lake. These purchases have been driven by the perceived opportunity to add low-cost reserves both from step-out drilling and the remediation of existing wells. In the case of Nomeco, the acquisition was a logical step in its expanding Antrim position as stated in its 1994 annual report:

"Two acquisitions of producing Antrim projects were made during the year. Total reserves added by the acquisitions were 9.4 bcf, involving interests in over 100 producing wells ... To date CMS Nomeco has participated in drilling or acquired interests in approximately 1,047 Antrim wells throughout the trend ... CMS Nomeco has invested a total of $105 million in Antrim projects and added 180 bcf of gas to the company's reserve base since becoming involved in the Antrim play. The program has yielded an overall finding cost of $3.50 per net equivalent barrel."

Another key change is the entry of several E&P majors into the Antrim play. Of particular note are the projects by Shell in Montmorency County and Amoco in Ingham County, at the southern edge of the basin, that have greatly expanded the potential productive trend. Other operators have received funding for new projects from outside sources such as Enron Capital and Trade. This suggests that industry has faith that this gas play can thrive and significantly expand without the tax credit.

Annual gas production from the Antrim shale has risen dramatically, from 26 bcf in 1990 to 128 bcf in 1994 (Table 1 [13584 bytes]). Daily gas production has increased from 30 MMcfd in 1990 to 350 MMcfd at the end of 1994 (Fig. 3 [23871 bytes]). Increased gas production is due to a combination of more wells and increased productivity per well. One incentive for this growth has been the stronger gas prices in the Michigan basin. From mid-1992 through mid-1994, the gas price was consistently more than $2/Mcf. An obstacle to the play that existed until 1995 has been the ability to fully transport the produced gas out of the basin. Traditionally, operators injected their gas into the Wet Header System, a regional trunkline through the play area originally installed for Niagaran gas production. But as Antrim production swelled, this pipeline soon reached capacity and became a bottleneck. Recently, a loop has been installed that has increased pipeline capacity out of the basin.

Contrary to those that called the Antrim shale a tax subsidy play that would die with the end of tax credits, an estimated 1,175 Antrim wells have been drilled and completed since 1992, the year tax credits stopped. The initial wells were drilled primarily in Antrim, Otsego, and Montmorency counties in the traditional play areas, with the more recent wells drilled in Alpena, Alcona, Manistee, and Oscoda counties to the east and south. Over 30 wells, drilled by Amoco, are in Ingham County at the southern edge of the basin. At the end of September 1995, 4,600 wells are producing 440 MMcfd, with 1995 annual production expected to reach 160 bcf from 4,600 wells.

In spite of the impressive number of wells drilled, only a small portion of the gas shales in the Michigan basin has been developed to date. Expanding this gas play will require technologies for identifying the highly naturally fractured, gas rich areas of this basin and continued improvements in linking this naturally fractured, multi-horizon reservoir to the wellbore.

Fort Worth basin

A more recent addition to the gas shale play is the Barnett shale in the Fort Worth basin of Texas. The first Barnett shale gas well was drilled in 1981 in Wise County by Mitchell Energy & Development. Eight years later, by the end of 1989, there were still only 52 gas shale wells drilled and completed in this basin. Spurred by improved characterization of these fractured, organically rich, self-sourcing reservoirs and advances in well location and completion practices, well drilling in this gas play has increased steadily in recent years.

Approximately 100 wells were completed during 1990-93, over 40 wells were completed in 1994, and a higher drilling pace exists in 1995.

Much of the early activity centered in eastern Wise and western Denton counties, although the organically rich shale is deposited widely throughout the Fort Worth basin. The current wells are drilled to 7,000-8,000 ft and completed in the Mississippian-age marine shelf shales. Based on resource characterization work by Mitchell and the Gas Research Institute, the shales have a porosity of 4-6% and a total organic carbon content of 5 wt %. The shale matrix is extremely tight, one micro-darcy or less, thus natural fractures are the key to productive wells. Because the shale has considerable thickness, the resource concentration is high, on the order of 30 bcf/sq mile (640 acres). Current well spacings of two wells per square mile provide opportunities for considerable future infill development.

Wells in the Barnett shale are drilled and completed to avoid damaging the formation and then massively fractured using 1.2-1.7 million lb of sand, depending on the thickness and number of pay zones. Treatment sizes and completion intervals have progressively increased since the initial wells were drilled in the early 1980s. To experiment with an alternative completion method, Mitchell (with support from GRI) drilled a 2,600 ft horizontal well to examine the applicability of this technology to the Barnett shale. While technically successful and valuable for characterizing the Barnett shale reservoir, the horizontal well contacted only a small fraction of the total pay and gas production was low. Thus, vertical wells located in areas of productive natural fractures and massively stimulated remain the preferred technology. Vertical wells in this area cost on the order of $500,000 to drill and complete plus about $200,000 to fracture. Surface equipment and infrastructure add to these costs.

Mitchell has been the leading operator in this gas play, accounting for 187 of the 203 wells tabulated. Other operators include Anadarko Petroleum Corp., Lakota Energy, Ryder Scott, and Winchester Exploration. Mitchell's commitment to this gas play is set forth in its last annual report that states:

"On top of a 160% replacement rate in fiscal 1994, 171% of the gas produced in fiscal 1995 was replaced. North Texas-and particularly the Barnett shale-played an important role in both the gas reserves gain and the finding cost improvement ... By finding naturally fractured "sweet spots" and further enhancing their productive capability with hydraulic fracturing, the company has made the Barnett an important new source of gas from our half-million-acre Fort Worth basin leaseholdings. During fiscal 1995, most of the 70 wells completed in North Texas were in that formation."

Gas production from the Barnett shales has increased steadily, reaching nearly 50 MMcfd at the end of 1994 from below 10 MMcfd in 1990 (Fig. 4 [23265 bytes]). With the strong well drilling since the end of the tax credits (in 1992), more than half the gas production from the Barnett shales is from post-1992 wells that do not qualify for the tax credit. This gas play also benefitted from a favorable gas price contract between Mitchell and Natural Gas Pipeline Co. that has since been renegotiated.

With the end of the favorable gas contract (and the earlier expiration of the tax credits), the Barnett shale gas play now must compete in the regional gas price market with otber gas supply sources. What is enabling this play to compete is the progressive improvement seen in the gas production rates and reserves per well (Table 9 [13839 bytes]).

  • The early Barnett shale wells, the 52 wells drilled in 1981-1989, had initial production rates of 300-400 Mcfd and expected recoveries of about 0.5 bcf/well.

  • With improved zone selection and stimulation, the subsequent 90 wells, brought on line in 1990 through 1993, showed substantially improved initial rates of 600 Mcfd These wells are expected to recover 0.5 bcf/well in 5 years and ultimately recover over 1 bcf/well.

  • The most recent wells, the 61 we1ls drilIed in 1994 and early 1995, show improved initial gas production rates and a less steep decline in the production rate and thus have higher potential for strong reserves per well.

    Debottlenecking of the gathering and compression system, installing improved well operations, particularly with respect to lifting of formation water and selection of back pressure, and advanced methods for placing new wells into naturally fractured settings should further improve the performance of the Barnett shale gas play. One example of this is the recently completed Barnett shale well, the Stella Young Gas Unit A6, which tested at more than 5 MMcfd.

Recently, Mitchell an- nounced plans to increase its well drilling and add 25-30 MMcfd of gas production from this gas play in the next 12 months.

Illinois basin

The most recent addition to the gas shale play is the Devonian and Mississippian-age New Albany shale of the Illinois basin, located in Illinois, southwestern Indiana, and western Kentucky.

Gas has been produced from the New Albany for more than 100 years, prirnarily in western Kentucky and southwestern Indiana, although historical production records are virtually non-existent. Technology developed in the Michigan basin Antrim has helped stimulate the interest in this area, and it appears that the New Albany shale has the potential to become the next big gas shale play.

In many ways analogous to the Antrim, typica1 drilling depths in the Illinois basin are around 1,000 ft and target the 100+ ft New Albany shale. Gas contents appear to be somewhat lower than in the Antrim, with similar water production characteristics. While the best stimulation approach for this formation is still uncertain at this time, multiple-stage hydraulic fracturing appears to be effective.

As of late 1995, 25 New Albany test wells have been driIled during the year in southern Indiana. The list of operators includes Four Sevens Oil, Mercury Exploration, Deka Exploration, and Minihan Oil & Gas. While spread throughout 10 counties, nine of the wells are located in Harrison County.

Successful technologies

A number of key technologies, currently developed or under development by a combination of industry and GRI, DOE, and the National Aeronautics & Space Administration, have enabled the gas shales play to survive the end of tax credits and continue during the past 3 years of low gas prices. Of particular note and value have been the technologies of enhanced understanding of reservoir dynamics, new methods for natural fracture detection, improved stimulation design, and well remediation.

  • Understanding of reservoir dynamics. An improved understanding of reservoir dynamics leading to better well and field development strategies has helped maintain development in all gas shales basins. Gas shale reservoirs are by nature highly complex, being triple porosity/dual permeability, with gas storage and release governed by desorption, and frequently having two-phase (gas and water) flow. Understanding these mechanisms is key to successful project development. Sophisticated reservoir simulation models have evolved to improve operator understanding of how best to develop these complex gas resources. One such model is ARI's state-of-the-art 3-D Comet2 reservoir simulator, which captures the unique properties of gas shale reservoirs. Comet2 has enabled operators to optimize well spacings, identify infill drilling opportunities, and even develop different zones in a multicompletion setting on different drainage spacings.

    As important as having a model to realistically simulate reservoir performance is the ability to effectively characterize the shale reservoir for well completion decisions. There have been many advancements along these lines the past several years, including the development of a gas content log that provides foot-by-foot estimates of gas content to help identify potentially productive horizons, sonic and resistivity-type wireline logging tools to identify intervals of natural fracturing, and core analysis methods to more precisely measure gas desorption and flow. These techniques are but a few of the important reservoir characterization advancements over the past few years.

  • Natural fracture detection. The existence of natural fractures in a shale reservoir is critical for commercial production. Natural fractures provide the conduit through which gas and water can flow from the tight shale matrix to the wellbore. Consequently, techniques to identify areas of intense natural fracturing in advance of drilling can make the difference between project success and failure. One technique, being demonstrated via an existing R&D contract between ARI and NASA, involves a combination of basin modeling, gravity surveys, and satellite imagery to predict areas of intense natural fracturing before drilling. This method has proven to be effective at locating areas of enhanced formation permeability in a number of gas shales, coalbed methane, and tight gas plays in the U.S. as well as overseas. Other techniques, such as shear-wave seismic utilized by Mitchell in the Barnett shale, have also been shown to be effective for locating areas of enhanced shale permeability in advance of drilling.

  • Improved stimulation designs. Improved well stimulation has helped sustain new well drilling in gas shales during this era of low gas prices. The underlying theme has been to better match well stimulation to inherent reservoir characteristics. One example of such applied technology is the use of liquid CO2 as a fracturing fluid in the Appalachian basin. The shale reservoirs in this basin have been known to be dry for many years, yet fracturing fluids such as nitrogen foam (containing at least 30% gelled water) have been typically used for hydraulic fracturing. The potential for formation damage resulting from the gelled water imbibing into and damaging the permeability of gas flow in the shale has been a concern to many operators. To address this problem, the DOE recently performed a field R&D study to investigate the potential application of nondamaging fracturing fluids in Appalachian basin gas shale wells. The results showed that straight-nitrogen treatments, without liquids or proppant, provided on average twice the production rate during the first 9 months compared to nitrogen-foam treatments. These tests also indicated that non-aqueous liquid CO2 treatments (with proppant) result in a four fold increase in production, on average. While these techniques have yet to be widely applied by basin operators, they hold promise for improving the productivity of gas shale wells in the Appalachian basin, which should lead to better well economics and drilling activity.

    Improvements in stimulation technology have also had a favorable impact in the Antrim shale. Here, the focus has been on how to best stimulate multiple pay horizons. At a GRI-sponsored field R&D experiment, performed by ARI in cooperation with Ward Lake Drilling, it was shown that perforating a two-stage stimulation to contact the two target shales (the Lachine and Norwood) could substantially improve production over traditional single-stage treatments. Now, some operators are performing up to three treatments per well-two in the thicker Lachine and one in the Norwood-to further improve well performance.

  • Well remediation. The area receiving the greatest focus in gas shales the past 3 years has been enhancement of gas production from existing fields. In the rush to drill and complete wells by yearend 1992 (to qualify for the tax credit), less-than-ideal completion and stimulation procedures were frequently employed that led to sub-optimum well performance.

    Operators have found that remediation of these older, sub-optimum wells can add substantial low-cost, low-risk reserves. Since remediation involves reworking older wells (drilled before yearend 1992), the incremental gas preduction from remediation often has the added benefit of qualifying for the tax credit.

    Most of the technologies utilized for successful remediation have been demonstrated in the Antrirn shale of the Michigan basin, including-well performance diagnostics, restimulation, and improved production practices. In one example, at Nomeco's Bagley East Project, a combination of technologies in all three areas were successfully utilized by ARI as part of a GRI funded R&D field experiment. Pressure transient testing was employed to identify a particularly attractive remediation candidate well, two-stage hydraulic fracturing was employed to effectively restimulate the two producing shale horizons (which were originally stimulated with a single treatment), and a downhole pump was installed to replace the reverse gas-lift dewatering system to lower flowing bottomhole pressure. Each of these techniques proved remarkably successful, together improving production immediately in the test well from 62 Mcfd to 255 Mcfd, and 9 months later to 410 Mcfd (Fig. 5 [22250 bytes]).

    Additional R&D is being focused at lower-cost approaches for identifying remediation candidate wells, evaluating Upper Antrim up-hole recompletion potential, and developing cost-effective, nondamaging ap- proaches to deepening previously drilled wells into the underlying Norwood formation.

    Advances in technology such as those described above have enabled the established gas shales plays to further develop and new plays to evolve, even after the tax credit expired. As a result, gas shales continue to represent a substantial opportunity for adding low cost gas reserves.

The Authors

Scott Reeves is a vice-president with Advanced Resources International Inc. in Arlington, Va. His training and experience are in the area of petroleum engineering and project management for both commercial and research applications. He has been involved in the oil and gas industry since 1984, managing and supervising exploratory, development, and remediation projects within the U.S. and overseas. He holds a petroleum engineering degree from Texas A&M University.

Vello Kuuskraa is president and cofounder of Advanced Resources International, which specializes in geological and engineering services to the oil and gas industry. He has more than 20 years' experience in energy resources development, technology, and economics. Before founding ARI, he was chairman of ICF Resources and cofounder of Lewin & Associates Inc. He received an MBA from the Wharton Graduate School of the University of Pennsylvania and a BS in mathematics/economics from North Carolina State Universitiy.

David Hill is a principal product manager, supply products with the Gas Research Institute. He is responsible for commercialization of products and process. Before this he was manager of emerging resources with focus on new and emerging resources such as coalbed methane and gas shales. Before joining GRI, he was a field engineer for Halliburton Services in the eastern U.S. He earned a BS in petroleum engineering from Marietta College.

Copyright 1996 Oil & Gas Journal. All Rights Reserved.