Lesley W. Evans,1 Barbara F. Keusch, 1 Stephen D. Sturm, 1 and William J. Clark1
Search and Discovery Article # 20006 (2001)
*Adapted for online presentation from poster session presented at the AAPG Convention, Denver, CO, June 2001. Please refer to: Topical Report GRI-00/0026, entitled “Multi-Disciplinary Analysis of Tight Gas Sandstone Reservoirs, Almond Formation, Siberia Ridge Field, Greater Green River Basin,” prepared for Gas Research Institute, Exploration and Production Business Unit, January, 2000, by Schlumberger Holditch-Reservoir Technologies, Inc., Denver, CO.
1Schlumberger Holditch-Reservoir Technologies, Denver, CO (www.slb.com) ([email protected]). Acknowledgments are extended to GTI (formerly Gas Research Institute), BP-Amoco, The Discovery Group, GeoQuest ICS, and Schlumberger Holditch-Reservoir Technologies. ** Mark of Schlumberger.
Editorial Note: This article, which is highly graphic (or visual) in design, is presented as: (1) nine posters, with (a) each represented in JPEG by a small, low-resolution image map of the original; each illustration or section of text on each poster is accessible for viewing at screen scale (higher resolution) by locating the cursor over the part of interest before clicking; and (b) each represented by a PDF image, which contains the usual enlargement capabilities; and (2) searchable HTML text with figure captions linked to corresponding illustrations with descriptions.
Users without high-speed internet access to this article may experience significant delay in downloading some of these illustrations due to their sizes.
The Siberia Ridge Field Reservoir Characterization study is based on full-field geoscience, petrophysical and engineering analyses including the results of a cooperative research well. This Gas Research Institute funded study provides insight into mechanisms controlling gas production in Siberia Ridge Field, southwestern Wyoming. The purpose of the study was to characterize the Almond Formation (Cretaceous) to better understand controls on productivity and to compile this information for analogous tight-gas sandstone reservoirs.
Geological facies analysis, petrophysical data, and seismic attribute data were mapped with production data to determine productive trends. Coherency analysis was used to determine the location of significant linear features. The combination of depositional, petrophysical, and structural data revealed areas of better petrophysical properties that generally indicated better production. This was apparent even within a small range of reservoir quality.
With average reservoir porosity ranging from eight to ten percent and matrix permeability in the microdarcy range, the role of natural fractures to productivity was of particular interest. FMI** and core data were used to characterize the natural fracture system. Natural fracture density was influenced more by lithology than lineament proximity or increased intersections resulting from wellbore deviation. Rather than providing increased conductivity to natural gas, natural fractures were found to provide increased relative permeability to deeper Almond water.
Specific completion practices were evaluated with geological and petrophysical data to determine relationships to gas production. Results indicated that completion practices were critical to well performance.
Enhanced reservoir understanding can improve drilling, completion and production practices, ultimately affecting well economics by decreasing risk and increasing recoverable reserves.
The following are objectives and scope of the project (Figure 1.1):
Integrate results of seismic and reservoir property mapping to identify discrete relationships:
· Better understand the interplay between matrix and fractures with regard to reservoir performance.
· Understand subsurface natural fracture density and relate this to improved gas recovery and lineament proximity.
Compare these findings with historical completion practices to determine effective methods to improve gas recoveries.
Reservoir Performance is the summation of the Reservoir and the Production and Engineering (Figure 1.2). In this study, key features related to the reservoir are the matrix (facies, especially lineament / channel focusing and coal distribution) and the fractures, especially as they relate to lithology and lineament proximity. Reservoir components, perforation practices, and stimulation practices are key elements of production and engineering practices.
Tasks to perform include the following (Figure 1.3):
Determine areas of accommodation (faults, linear features, paleo-lows, interval thickness).
Integrate existing data and maps to determine favorable depositional/petrophysical facies.
Balance objectives with acreage and EIS issues.
Determine location and drill.
“Short Drill” ALM bar - ALM400.
“Long Drill” ALM bar - ALM700.
Run and evaluate mud and open hole logs.
Option to set intermediate casing.
Run mud and open hole logs include FMI**.
Select intervals to perforate (perforation density not less than 4spf ).
Evaluate lower Main Almond interval for Sw and fractures to mitigate water risk.
Select stimulation grouping (less than 100 feet gross interval).
Run Plt’s and obtain pressure data periodically to evaluate performance.
The Siberia Ridge Reservoir Characterization study area is located on the north flank of the Wamsutter Arch, in the Great Divide Basin of southwestern Wyoming (T21-22N and R94-95W) (Figure 2.1a). The field area is structurally simple, dipping to the northeast at approximately 1° to 1.5°, or ~150 ft/mile (Figure 2.1b). In 1975 the discovery well was completed in the Upper Cretaceous (Campanian) Almond Formation of the Mesaverde Group with an initial potential of 4,204 MCFGPD. Gas and gas condensate (API 50°) are primarily produced from the laterally continuous Upper Almond “bar” sandstone at the top of the formation (Figure 2.2), with supplemental production from the underlying “Main Almond” sandstones.
Productivity varies significantly throughout the field. Detailed reservoir analysis indicated that a complex combination of depositional matrix characteristics, sandbody dimension and connectivity, mechanical attributes (fracturing), and coal (source) are highly variable within a restricted geographic domain, but all affect production.
In the Washakie Basin, Almond Formation thickness varies from 250 to greater than 500 feet. These variations in lithofacies are thought to be due to syndepositional movement along basement block faults (Martinsen et al., 1995).
The Almond is a major transgressive sequence, composed of smaller transgressive-regressive cycles that have been divided into the Upper and Main Almond zones (Figure 2.3). The Upper Almond is composed of marine sandstones, whereas the Main Almond consists of brackish to nonmarine interbedded sandstones, siltstones, shales and coals.
The underlying Ericson Sandstone (Figures 2.2 and 2.3) is predominantly nonmarine, composed of amalgamated fluvial sandstones interbedded with floodplain deposits. Within the area of the Siberia Ridge Field, the Ericson-Almond contact is vaguely defined.
The Almond Formation can be divided into two genetic units, the Main and Upper Almond (Figures 3.1 and 3.2). The division between these two units is typically defined by the initial occurrence of a correlatable transgressive marine shale above the continental Main Almond interval.
The Almond 100 through Almond 700, or Main Almond, is approximately 450 ft thick and is composed of 40-70 ft ‘depositional sequences’ dominated by lenticular, tidal flat and tidal channel sandstones encased within bayfill and estuarine shales. Individual sequences are bounded by continuous coal beds and carbonaceous shales. These sequences become increasingly marine upwards. Sandstones in the Main Almond are dominated by tidal facies which are highly compartmentalized and have limited connectivity. Drainage area or radius generally falls below well spacing.
The composite sandstone at the top of the sequence, commonly referred to as the Almond ‘bar,’ is the primary pay zone. Within Siberia Ridge Field, the Upper Almond is 30-50 ft thick. This sandstone is composed of amalgamated shoreface and tidal channel sandstones and is laterally continuous. Permeability ranges from 0.001 to < 0.1md, average kh is 0.263md-ft, and porosity ranges from 6-12% (average 10.3%) (Figures 3.3 and 3.4).
Detailed sedimentological, stratigraphic and petrophysical characterization identified two distinct, northeast-trending tidal channel complexes that dissect shoreface deposits in the field (Figure 3.5). Comparative analysis of reservoir characteristics in shoreface and tidal channel sandstones indicates that grain size, volume of dolomite cement, and secondary porosity are primarily responsible for differences in reservoir quality (Figure 3.6).
Petrographic and petrophysical data indicate that tidal channel sandstones are relatively coarser grained, are silica and clay cemented, and are subjected to dissolution of unstable rock fragments resulting in secondary porosity. In contrast, shoreface sandstones are finer grained and contain high volumes of dolomite cement.
Linear features were initially identified with seismic data. Figures 4.1 and 4.2 compare the derivative of the Madison-Almond isochron map (Figure 4.1) and the seismic dip map (Figure 4.2). In Figure 4.1, the areas of red and yellow denote regions of greater flexure within the Madison-Almond time interval. The seismic dip map is a display of the magnitude of the time dip of the autopicked Almond seismic event. Lighter shades of gray denote areas of relatively flat dip, whereas darker shades indicate steeper structural dip.
The three most significant features are the en echelon northeast-trending linear features: the SRU and 5-2 which roughly correspond to the Ransome lineament as drawn on Figure 4.3 and the Wamsutter lineaments. The northwest-trending features (N1-N10, S1-S7) terminate abruptly at the 5-2 lineament and cannot be verified from wellbore data; however, the orientation of these segmented linear features is also identified on the modified Digital Elevation Model (Figure 4.4).
The yellow circles on the seismic dip map (Figure 4.2) indicate the locations of wells with described cores. The three key fracture wells are:
•SRU #27-4 (FMI**, DSI), 1380 ft from Wamsutter Lineament, 313 ft from S5.
•SRU #5-2 (FMI**, DSI and Core, deviated wellbore), 814 ft from 5-2. Lineament at surface; intersects 5-2 Lineament in the ALM400 interval.
•SRU #5 (Core), 3,000 ft from nearest lineament.
Analysis of borehole fracture data indicates that fractures in Siberia Ridge Field are systematic regional extension fractures. They are mineralized by quartz druze and kaolinite; dip is near vertical; and strike is between N60°E and N70°E (Figure 4.5).
Fracture density in the deviated (45º) Siberia Ridge #5-2 Almond bar through Almond 300 intervals (Figure 4.6) is greater than in the vertical wells (Figure 4.7). However, there is a remarkable change in fracture frequency at the Almond 400 interval, where the average spacing decreases to one fracture per 1.3 feet. This dramatic increase in fracture density is due to an increase in net sandstone and to the intersection with the #5-2 lineament (Figure 4.8)
Siberia Ridge vertical fracture density ranges from one fracture every 5.5 feet to less than 1 fracture per foot. Fracture density increases with depth in each of the key Siberia Ridge wells below the Almond 300 interval (Figures 4.6, 4.7, 5.1). The apparent increase in fracture density with depth (Figure 5.2) may be explained several ways: increased sand penetration (Figure 5.3), changes in mechanical properties of the sandstones, and closer proximity to local increases in stress. Increased fracture density in the Siberia Ridge #5-2 and Siberia Ridge #27-4 wellbores corresponds to penetration of thicker, lower Main Almond and Ericson sandstones. The Siberia Ridge #5-2 is also less sandy than the SRU #27-4 (by a factor of two). The Siberia Ridge #5-2 was drilled to take advantage of a regional structural lineament, whereas the Siberia Ridge #27-4 was not. The difference in fracture density (normalized for lithology and wellbore deviation) between these wells, however, is only a factor of 1.4, suggesting that lithology strongly influences the presence of natural fractures in these wellbores.
Investigation of the wellbore geometry shows that the fracture spacing reported for the Siberia Ridge #5-2 is in the Z direction, whereas those reported for the Siberia Ridge #27-4 and Siberia Ridge #5 are in the Y direction (Figure 5.4). Even though the Siberia Ridge #5-2 is deviated, the average fracture spacing, as determined from FMI**, is similar to the Siberia Ridge #27-4 (Figure 5.2).
To understand the effect of drilling near and/or through the 5-2 Lineament, the fracture spacings are compared on the same plane. Siberia Ridge #5-2’s fracture density converted to fracture spacing, DX, equals 4 inches, whereas fracture density for the vertical wells converted to fracture spacing, DY, is 26.8 inches.
What is the effect of wellbore deviation? Utilizing the geometric relationships obtained from these figures and equations, the spacings in each plane are reported in Table 5.1.
The geometrical difference between DZ and DY spacing is a factor of 4.7, and the fracture density difference between the wells normalized for lithological effects is 6.7. This leaves just a factor of 1.4 as the increase in natural fracturing possibly due to fault proximity. This result is surprising, since it suggests that wellbore deviation impacted fracture spacing more than lineament proximity in the Siberia Ridge #5-2.
In summary, wellbore deviation impacts fracture density by a factor of 4.7; lineament proximity impacts fracture density by a factor of 1.4
Fracture permeability is inversely proportional (both numerically and linearly) to the distance between fractures. Fracture permeability is also proportional to the cube of the fracture width. From a mathematical standpoint (Table 5.2), the fracture width effects fracture permeability more than fracture spacing (Figures 5.5, 5.6, and 5.7).
The bar graph (Figure 5.8) of mean fracture permeability (k2, parallel planes) shows mean k2 permeabilities of 5 md to 225 md. Fracture permeability is lowest in both wells (less than 10 md) in the Almond Bar to Almond 200 intervals, where fracture spacing is also the greatest. Fracture permeability increases to more than 50 md in the Almond 300 interval in the Siberia Ridge #5-2, and the Almond 400 interval in the Siberia Ridge #27-4. This permeability increase corresponds to the increase in fracture density.
There is tremendous fracture permeability (225 md) in the Siberia Ridge #5-2, Almond 500 interval. Since fracture spacing is the same as the Siberia Ridge #27-4 in this interval, why is fracture permeability so high?
A glance at the distribution of fracture apertures (Figure 5.9) answers this question. The very large apertures in this interval are masked by the interval mean. For this reason, fracture data needs to be manipulated and viewed a number of different ways. Averaged data allows for ease of comparison between wells or intervals, but point to point fracture data are necessary to compute discrete fracture permeability values.
Gas production in Siberia Ridge Field is quite variable, with estimated ultimate recoveries (EUR) averaging 1.8 BCF and ranging from less than 0.5 to nearly 20 BCF (Figure 6.1). Are productive sweet spots controlled by increased natural fracturing, better matrix quality, completions practices, or by a combination of these factors?
Most wells are completed in both the Almond Bar and Main Almond. Almond gas production is usually commingled, making accurate assessment of interval productivity difficult.
Initial production rates for the Main Almond are higher and decline more rapidly than production from the Almond Bar (Figure 6.2). This is due to the areally limited size of Main Almond reservoirs. Production logs (Figures 6.3 and 6.4) support this geologic interpretation, which is also observed in the decline curves for Main Almond wells. However, water production also becomes an issue when perforating and producing the lower Main Almond Formation.
Perforating the lower Main Almond increases exposure to high water production. Many of the Almond 600 and Almond 700 perforations are plugged back, and few wells are perforated below the Almond 400 interval (Figure 6.5). Historically, water saturation is calculated assuming a long, single hydrocarbon column; yet water often produces from the lower Main Almond and Ericson intervals. Conventional log analysis has not yet adequately determined why these reservoirs produce water, given an increasing resistivity profile with depth. Porosity is still developed in the lower Main Almond intervals, and log analysis generally indicates better gas saturation in these intervals than in the Upper Almond Bar. Variations in Rw have been proposed to explain the apparent saturation change, and a radical change in Rw from 40,000 ppm in the bar to 5,000 ppm would be needed to calculate a “wet” zone in the lower Almond intervals. Two things are needed to support a change in Rw; one would be a fresh water source from the coals (probably during gas migration), the other would be vertical compartmentalization within the Almond Formation. More work to understand the mechanics of coal gas generation, water migration, and vertical compartmentalization in the lower Main Almond is needed. The combination of high natural fracture density (from this study) and a drastic change in rock type from the Upper Almond bar to the Ericson Formation can account for the difficulty in determining “wet” zones in this field. An approach utilizing capillary pressure techniques may be better suited to determine saturation.
What is the impact of different completion technologies? The graphs in Figures 6.6 and 6.7 demonstrate that higher perforation density (Figure 6.6), limiting the gross stimulated interval to under 100 feet (Figure 6.7), and utilizing discrete or point sources treatment methods improve gas recovery.
An increase in initial and long term productivity is observed in some wells intersecting linear features. The level of increased productivity may depend on the stratigraphic intersection (Table 6.1). Intersection in the Upper Almond interval potentially exposes a larger productive area than the smaller lenticular reservoirs of the Main Almond intervals. Linear feature intersection at or below the Almond 400, where fracture density normally increases significantly, also increases exposure to potential water sources.
For wells perforated only in the Upper Almond-interval, simple linear regression between EUR and distance to linear feature results in a R2 value of 0.28; whereas a regression between EUR and kh results in a R2value of 0.64. A multilinear regression of EUR with kh and distance to linear feature results in a R2 value of 0.744, suggesting that the distance to the linear feature improves the overall correlation. This finding supports the idea that kh is more significant than distance to linear feature in long term productivity but also indicates that distance to linear feature is a factor in long term productivity.
Figure 7.1. Comparison of the Upper Almond facies and production map (a) to the Almond-Madison isochron (b). The drainage divide separating the two northeast-trending channel complexes is thought to be a syndepositional feature related to inherited Paleozoic structural features. The SRU Lineament defines the hinge of the paleostructure and Upper Almond drainage divide. The 5-2 Lineament forms the approximate southern boundary of the northern channel complex. Bubbles of cumulative gas production (normalized for the 1st 180 days) are displayed on the facies map.
Figure 7.2. Facies control on production: 27 Upper Quartile Wells (blue circles) produce 50% of the total gas (EUR); 84% of these wells are within the tidal channel facies (orange area) of the Upper Almond.
Figure 7.3. Combinations of matrix and mechanical (natural fracturing) attributes are compared to initial 180 days of production. These areas of upscaled attributes (kh, coal, fractures) account for approximately 60% of the Almond gas.
1. Linear features delineate paleo-lows, where better reservoir quality sands were deposited in channels.
2. Mapped areas of higher matrix quality are generally areas of higher productivity.
3. Roughly 90% of produced gas is from the Almond Bar.
Natural Fractures (Figure 7.3)
1. Fracture density depends mostly on lithology.
2. Wellbore deviation effectively increases fracture density (permeability) more than lineament proximity.
3. Water production in the SRU #27-4 and SRU #5-2 is from highly fractured intervals.
1. Decline curves from Upper Almond Bar wells are flatter than Main Almond reservoirs due to the difference in reservoir size.
2. Tidal channel reservoirs (Figure 7.2) have a better EUR than shoreface reservoirs due to matrix quality.
3. Hydraulic stimulations stages should access less than 100 feet per treatment. Single reservoirs should be targeted.
Economic success is measured by improved well recovery and reduced cost through site selection, pay determination, natural fracture characterization, completion optimization, and recompletion potential. These are achieved by:
Designing appropriate 3-D seismic for target interval.
Determining areas of accommodation (faults, linear features, paleo paleo-lows, interval thickness).
Integrating with existing data and maps to determine favorable depositional and depositional and petrophysical facies.
• Targeting channel areas of high matrix quality.
• Intersecting linear features in the Almond bar.
• Avoiding lower Main Almond sandstones unless precautions are taken for overpressure. Recalculate Rw and Sw for proper formation interpretation.
Running mud and open hole logs including FMI** for fracture interpretation.
Drilling non-vertical wellbores.
Utilizing staged hydraulic stimulation techniques with a minimum perforation density of 4 shots per foot.
Implementing the information presented in this study will increase asset value through improved well recovery, higher initial rates (IP), and a decrease in drilling and operational costs. This is accomplished through better infill drilling location selection and advantageous completion practices. As a result, average infill locations will recover greater than .5 to 1.5 BCF per well over historical Almond well completions.
Aguilera, R., 1995, Naturally fractured reservoirs, Second Edition: Tulsa, PennWell Publishing Company, 521p.
Gomez, E., Horne, J., Forster, J., Sturm, S., 1995, Stratigraphic framework and facies distribution of the Frontier Formation, Moxa Arch, Consortium for emerging gas resources in the Greater Green River Basin, in Geological characterizations of the Cretaceous gas plays in the Greater Green River Basin: GRI 95/0450, p 115.
Jaworowski, Cheryl, and Robert Simon, 1995, Relating fractures in the eastern Greater Green River Basin to oil and gas fields: integrating field measurements, linear features, and digital geologic data, in Rsources of Wyoming: WGA Guidebook, p. 61-75.
Martinsen, Randi S., Glen E. Christiansen, Mark A. Olson, and Ronald C. Surdam,1995, Stratigraphy and lithofacies of the Almond Formation, Washakie and Great Divide basins, Wyoming, in Resources of Southwestern Wyoming: WGA Guidebook, p. 297-310.