Matthew E. Blauch, Jim J. Venditto, David E. McMechan
Halliburton Services
Duncan, Okla.
Richard K. Smith
Halliburton Services
Kingsport, Tenn.
Paul V. Hyde
Columbia Natural Resources Inc.
Charleston, W.Va.
Perry A. Harris
Halliburton Logging Services
Charleston, W.Va.
Experimental and novel techniques, tools, and data integration can provide cost-effective exploration and development strategies for producing long-term local low-cost gas from Devonian shales.
This first of a five-part series will discuss a test program in Roane County, W. Va. in which Columbia Natural Resources (CNR) and divisions of Halliburton Co. combined efforts to develop and test new technologies for producing more gas from tight gas reservoirs. These special technologies include:
- Core analysis and integrated CT (computerized tomography) imaging
- Open hole logging
- Downhole extensometer and microfracturing techniques
- New surfactants
- Mechanical rock properties logging, oriented perforating, rotational radioactive tracer logging, and 3-D fracture modeling used in hydraulic fracturing.
The test well was drilled in late 1990 and fracture stimulated in 1991. The poststimulation evaluation and well testing operations included:
- Developing a fundamental understanding of geologic characteristics and controls on production
- Evaluating tools for open hole logging in air-filled wells
- Obtaining hydraulic fracture and in situ stress orientation
- Imaging techniques for shale characterization
- Selecting perforation intervals and orientation for fracture stimulation
- Evaluating stimulation fluids and surfactants
- Improving stimulation practices.
GEOLOGY
The study well is in the Devonian shale of the Appalachian basin (Fig. 1).
Devonian shale reservoirs produce primarily natural gas and localized oil.
This widespread shale contains at least 409 tcf of gas in place in the Appalachian basin region and about 570 tcf in all of the U.S. Natural gas can be obtained from three types of production in the shale:
- Associated gas from oil wells
such as in the Burning Springs anti- cline
- Natural gas liquids from wet gas
wells
- Natural gas from dry gas wells (the most common production type).
In most of West Virginia, the shale contains large gas reserve potential in abundant thick layers of black, organically rich shale that primarily produces dry gas.
But in Burning Springs, the Devonian shale interval is more siltstone and sandstone stringers that often produce hydrocarbon liquids as well as gas.
Devonian shales exhibit very low porosity and low permeability. Published core porosity values range from 1 to 8%, and permeability ranges from 0.01 md to less than 0.1 nanodarcies.
The Devonian shale gas reservoir is composed of identifiable features such as porous lithologies, natural fracture systems, and gas adsorbing materials. Fracturing is primarily caused by thrusting and the reactivation of basement faults.
Organic richness, organic matter type, and thermal maturity vary widely across the Appalachian basin and play a key role in the generation and storage of hydrocarbons.
Characteristics of organic shale content, siltstone distribution, and thermal maturity also directly control the shale's capacity to adsorb and release gas.
Shale mineralogy is predominantly quartz, illite, and micas with small amounts of carbonates, feldspars, other clays, and heavy accessory minerals. Detrital illites and micas mainly control rock microfabric and vertical permeability anisotropy because illites and micas are aligned preferentially parallel to depositional bedding (Fig. 2).
The porosity contains bound water, free water, gas, oil, and kerogen and is characteristically dominated by sheet-like pores formed by the detrital rock fabric.
Because rock stress contrasts between productive intervals and potential stimulation barriers are plentiful, vertical fracture growth containment is an important consideration for hydraulic fracturing.
Historically, well performance has been inconsistent. The mechanisms accounting for anomalous production have eluded researchers because very little information has been systematically obtained or analyzed from a single Devonian shale well.
Gas reserve potential in the shale appears to be derived from the total effect of a series of low-grade reservoirs stacked in the shale section.
Where present, natural fractures (joints and thrust faults) are known to enhance production in the presence of porosity and organics.
The role of desorption, although poorly defined regarding contribution in a given shale reservoir, appears to influence gas recovery in the long term.
With proper stimulation, small incremental increases in initial production can significantly increase return on investment for both new wells and recompletions.
WELL COMPLETIONS
Historically, most early Devonian shale wells were:
- Developed with minimal application of available technology
- Drilled using cable tool drilling rigs
- Completed open hole
- Stimulated using nitroglycerine explosives.
Because nitroglycerine fractures were multidirectional, unpropped, and limited penetration, a relatively small degree of stimulation could be obtained.
Although production rates were low compared to wells using current stimulation technology, many wells drilled more than 50 years ago are still producing today.
Other stimulation methods were attempted after the introduction of hydraulic fracturing in the late 1950s. Most shale wells were cased and cemented through producing intervals. Water and sand were pumped down the casing and into the formation.
Crude oil and sand were also used in some cases. Water and crude oil fracs helped improve production more than the nitroglycerine. However, no chemicals, chemical additives, or energized fluids were used until the 1960s.
In the late 1970s, foam fracturing with water, a foaming agent, and nitrogen improved results but still did not attain expectations due to designed fracture geometry and size.
Currently, several different types of hydraulic fracturing have advantages and disadvantages.
Nitrogen fracture treatments with only gaseous nitrogen under pressure are usually less expensive than water or foam fracture treatments. Because no liquids are used, liquid removal and retention are not a problem. However, without proppant, spalling of the formation face is relied upon for maintaining fracture conductivity.
Typical treatment sizes range from 1.5 to 2.0 MMscf of nitrogen. Most treatments are pumped at a rate of 100 Mscf/min.
Proppant-laden treatments generally include a foam or gelled water carrier. Foams and gelled water have good sand carrying qualities.
The aqueous fluids should be recovered because if water is left in the formation, relative permeability to gas will be reduced significantly along with productive capacity.
Devonian shale reservoir pressure usually is insufficient to overcome capillary forces and produce back the fracturing fluids. To solve this problem, foam fluids are preferred. However, little consideration has been given to the effects of the foaming surfactants on gas production and load water recovery. This remains a primary concern for foam treatments.
Typical foam treatments are 40,000-60,000 lb of 20/40 sand pumped at a rate of 40 bbl/min. Included is about 6,000-10,000 gal of aqueous stimulation fluid per stage (most jobs are run in three stages). Sand concentrations range from 1.5 to 4.0 ppg downhole.
Now, most foam treatments are 75 quality foams with some treatments starting at 90 quality and decreasing down to 65 quality at higher sand concentrations. These designs leave less fluid behind in the reservoir and are becoming widely accepted by many operators in the Appalachian basin.
EMERGING TECHNOLOGIES
In the test well, several open hole logging techniques were used in the air-filled hole. Unlike conventional logs, a high-resolution induction (HRI) log reliably evaluated thin bedding and reservoir delineation in the shale (Fig. 3).
Identification of the dominance of millstones and thin bedding from an X-ray CT imaging of the core supported the use of the HRI log for reservoir evaluation (Fig. 4).
Geologic evaluation and review also supported the hypothesis that the distribution of thin-bed, depositionally controlled millstones within the Devonian shale reservoir may be a primary control on gas production. Identification of the Devonian shale strata as a reservoir containing small-scale heterogeneous sequences has shown potential as a tool or basis for selection of candidate wells within regionally and stratigraphically variable Devonian shale intervals.
Coring provided inferred hydraulic fracture orientation and in situ stress orientation prior to successfully performing a full-scale hydraulic fracture job. Also, coring and core studies added important information for geologic and reservoir description.
The inferred hydraulic fracture and in situ stress orientations were determined and confirmed by several independent techniques including the recently developed borehole extensometer. The N60E fracture orientation from the extensometer evaluation during a downhole microfrac treatment was confirmed independently by the CT imaging fracture orientation analysis (Fig. 5).
Subsequently, this information was used for perforation orientation and breakdown by incorporating a novel downhole directional gamma ray tool designed to measure the azimuthal distribution of downhole radioactive tracers. This was one of the first tests performed in the industry to evaluate and confirm the effects of fracture orientation.
In addition, extensometer data showed that pressure displaced the borehole anisotropically. The displacement corresponded to the subsequent fracture orientation. This observation has significant implications regarding mechanical rock properties of the Devonian shale and other types of reservoirs.
An experimental method was tested for calculating gas permeability as it relates to gas diffusively from a core placed in an on site pressure chamber apparatus. The calculated permeability and analytic solution corresponded favorably with core and well test values.
Contrary to previous ideas, experiments with xenon gas invasion using CT imaging demonstrated that gas adsorption/desorption may not be as significant in the gas storage mechanism of the Devonian shale. Based on the experimental core research, free gas porosity appears to be the primary production mechanism.
A series of six experimental microfracture treatments were made with nitrogen only or KCI (potassium cloride) water. Tests showed that nitrogen-only microfracture treatments provided reliable stress information. Values obtained from these microfracs corresponded well with the stress log.
As a result, the stress log did not require calibration and can be used with confidence for designing fracture stimulations.
Significant information was obtained regarding effects and response of stimulation fluids on gas flow through core samples in the low-permeability Devonian reservoir. Previous testing technology did not permit realistic evaluation in the laboratory under low reservoir pressure conditions.
In the joint project, special equipment evaluated the effects of stimulation fluids on gas recovery. Results may change the interpretation of the effects of stimulation fluids on gas production in the Devonian shale and similar tight, gas-bearing formations.
The experimental core testing has led to the development of a new surfactant system for foamed stimulation treatment.
With minifrac testing, postminifrac well testing, and the fracture stimulation treatment, a downhole acoustic log was run after a fracture treatment. Fracture design modeling and perforating strategies based on the stress log showed more significant stress heterogeneity than expected. Potential future fracture design improvements could incorporate this information.
Poststimulation analysis of flowback samples included the effect of the recently developed foaming surfactant on stimulation fluid flowback response. A characteristic chemical analysis response of load recovery and formation water influence indicated efficient load water recovery with the experimental surfactant system used in the study well.
Treatments in other nearby wells indicated improved flowback volumes and rates as well as enhanced initial production values. In addition, improvements in breakdown and tubing cleanout procedures were achieved.
RECOMMENDATIONS
The CNR/Halliburton joint project produced a series of recommendations for logging, coring, interpretation of inferred hydraulic fracture orientation, stimulation fluid selection, perforation breakdown procedures and breakdown acid selection, fracture stimulation design, follow-ups for completion optimization, and structural/depositional geological interpretations.
LOGGING
The recommended logging suite for air-drilled open hole logging includes:
- HRI log
- SLD (spectral lithodensity) log
- Sidewall neutron log
- Temperature log
- FWS (full-wave sonic) log.
The FWS log was run in other wells to verify the borehole stress profile log. The FWS log matched the micro and minifrac test but was not verified with the main frac.
To run the sonic log, the well must be filled with fluid to sufficiently cover the interval of interest. It may be possible to run this log in cased holes provided that the casing diameter is at least 5 in. and the primary cement job is satisfactory (at least 50% bond).
An acoustic borehole imaging log should also be considered if the well is filled with fluid. This log would probably produce better results in identifying natural fractures when a microfrac test has not been conducted in the open hole.
A tracerscan log should be run after each hydraulic fracture job to help verify the treatment's effectiveness. This requires tagging each fracture with a radioactive isotope.
It is recommended that no more than three different tracers be used in these operations. If fracturing treatments are performed down the casing, the full-size 3/8-in. diameter tracer tool could be run. This tool is more efficient in determining the location of tracer material because of its larger detector.
Oriented perforating in the direction of anticipated fracture propagation provides a significant advantage over conventional phased perforating. Results of this experiment showed that azimuthally oriented perforating could:
- Improve the effectiveness of the fracture treatment
- Decrease near-well bore tortuosity
- Reduce the possibility of early screenouts.
CORING
Quality assurance of the core orientation package should be confirmed before coring. This should include alignment of the orientation tool with the core barrel.
Where orientation is critical, a continuous core orientation system is suggested.
An experimental inner core barrel stabilization technique was used as an integral part of the shale coring operations to obtain quality core samples, subsequent CT scanning, and fracture orientation analysis.
The core stabilization is suggested for obtaining optimal quality in any inner-barrel coring operation performed in friable, fractured, or difficult formations where maximum return on the coring investment is desired.
FRACTURE ORIENTATION
Fracture orientation data obtained from CT scanning of cored microfracs and the borehole extensometer showed excellent agreement. Fracture orientation from these methods is preferred over other techniques.
With additional hydraulic orientation data from offset wells and comparison to wells located in various local structural geologic settings, future plans include confirming apparent regional structural geologic control over hydraulic fracture orientation.
Prefracture displacement data obtained from the experimental borehole extensometer indicated an anisotropic response of the shale to pressure. The response correlated well with well bore breakouts observed with borehole imaging logs and the borehole televiewer.
The borehole imaging logs and oriented borehole imaging techniques showed that breakout orientations can determine in situ stress and the inferred hydraulic fracture orientation.
STIMULATION FLUIDS
Stimulation fluids with about 2% KCI provide adequate shale stabilization as compared with other temporary clay control materials but do not prevent significant permeability reduction from rock/fluid contact.
A recently developed surfactant system consisting of a novel foaming surfactant is the preferred choice for future foamed stimulation work. Contrary to previous beliefs that high-concentration methanol fluids should be used to stimulate the shale, the use of methanol in Devonian shale is discouraged.
Methanol did not show a favorable permeability response in the shale core testing study. Additionally, methanol-based fluids have not shown improved well responses.
A foam generator is suggested for foam fracturing.
PERFORATION BREAKDOWN
A 15% acid pickling treatment containing an effective ferric iron reducing agent is recommended before perforation breakdown. The acid should be circulated back to surface to prevent potential formation damage from subsequent ferric iron hydroxide precipitation.
To prevent rust formation on the pipe following a pickling treatment, acid should be displaced with nitrogen rather than rig air.
Perforation breakdown fluids should consist of non-HCl fluids such as 10% organic acid with corrosion inhibitor containing an appropriate surfactant. A nonacid surfactant/KCI fluid can also be considered.
For optimal breakdown and subsequent hydraulic fracturing response, perforations should be aligned at 180 phasing in the inferred hydraulic fracture orientation.
In oriented perforations, an experimental rotational directional gamma ray log confirmed the hypothesis of near well bore hydraulic fracture alignment through azimuthal perforation orientation and confirmed the orientation of the hydraulic fracture.
FRACTURE STIMULATION
An apparent stress discontinuity near the bottom of the Lower Huron shale interval needs to be identified to avoid attempting a fracture stage that straddles the discontinuity. The discontinuity was in close proximity to the bottom of the upper stage in the study well.
Tracer logs indicated much less height growth than had been expected. Typical perforation spacing did not result in a single large connected fracture.
The displacement anisotropy observed by the borehole extensometer indicated an effective overall formation stiffness (resistance to deformation) that was lower than the stiffness of the rock matrix sample tested in the lab. This would result in wider hydraulic fractures at lower extension pressures with less tendency for fracture height growth.
The stress also showed significant variations (100-300 psi) over relatively short intervals. This variation provides a series of weak barriers to fracture height growth.
For optimal well cleanup, fracture treatments should be staged sequentially and flowed back to surface after each stage for improved load recovery and ultimate production.
Each operator should develop a standardized procedure to track the results of systematic changes in completion techniques applied. Long-term production tracking and production logging are important for effectively evaluating completion and stimulation strategies.
GEOLOGIC ANALYSIS
Siltstone deposition is believed to be a primary control for gas production. Identification of shale strata containing small scale heterogeneities can enhance selecting candidate wells in various internals. The HRI log is suggested for characterization and classification of thin siltstone bedding.
CT imaging data and quantitative statistics have shown potential for core/log correlation in defining log response in the shale formation. Identification of mineralization along distinct bedding laminae and partings in the shale as determined from CT imaging can provide additional criteria for selecting productive shale intervals.
Correlation to log response can assist in identifying the characteristic productive lithology obtained from cores.
An apparent stress discontinuity was identified within the shale package. Subsequent stimulation design strategies should focus on geologic characterization and regional identification of significant stress contrast within the stratigraphic section.
A method of log identification should also be developed for air-filled holes for geological characterization. A lower stress below the discontinuity would significantly impact hydraulic fracturing designs.
Because of hydrostatic considerations, a fluid-filled open hole stress log could not be run across the stress discontinuity identified in the study well.
ACKNOWLEDGMENTS
The authors would like to express their appreciation to Columbia Natural Resources and Halliburton Co. for permission to publish results of this work, and to all who contributed to this project.
BIBLIOGRAPHY
- Calvert, G., "Subsurface Structure of the Burning Springs Anticline, West Virginia-Evidence for a Two-Stage Structural Development," Appalachian Basin Industrial Associates, Vol. 4, 1983, pp. 147-164.
- Ettensohn, F.R., "The Catskill Delta Complex and the Acadian Orogeny: A Model," G.S.A. Spec. Paper, No. 201, 1985, pp. 39-49.
- Luffel, D.L., and Guidry, F., "Reservoir Rock Properties of Devonian Shale From Core and Log Analyses," Paper SCA 8910, Society of Core Analysts, New Orleans, 1989.
- Perry, W.J. Jr., "Mann Mountain Anticline: Western Limit of Detachment in South-Central West Virginia: Western Limits on Detachment and Related Structures in the Appalachian Foreland," U.S. DOE, Morgantown, W.Va., 1978, pp. 82-99.
- Dennison, J.M., "Stratigraphic Distribution of Decollements in the Appalachian Basin," Geological Society of America Abstracts With Programs, Vol. 19, No. 4, 1978, p. 167.
- Fodgers, J., "Mechanics of Appalachian Foreland Folding in Pennsylvania and West Virginia," AAPG, Vol. 47, No. 8, 1963,
- Wheeler, H.E,, "Post Sauk and Pre-Absaroka Paleozoic Stratigraphic Patterns in North America," American Association of Petroleum Geologists Bulletin, Vol. 47, 1963.
- Sweeney, J., "Oil and Gas Report and Maps of Wirt, Roane, and Calhoun Counties," West Virginia Geological and Economic Survey Bulletin, B-40, P. 102.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
