The Natih Formation in the Natih field is a heterogeneously fractured reservoir being developed by gas/oil gravity drainage. An understanding of the degree and orientation of the fracturing is essential for the optimum development of the reservoir. In order to better understand the fracturing in the reservoir, conventional 3-D (compressional) and limited 9C3D (nine-component, three-dimensional) seismic surveys were made of the Natih field. A revised fault/fracture model was developed from the conventional 3-D survey results in which ‘domains’ of similar fault/fracture character have been defined. Comparison with well-production data indicates that the domains associated with strike-slip or rotational movements (interpreted as being more fractured than adjacent areas) are zones of significantly higher productivity. The 9C3D seismic survey results also show areas or ‘domains’ of variable shear-wave time splitting. A comparison of the fault/fracture domains with the shear-wave time-splitting domains shows a close relationship in which areas of highest shear-wave time splitting coincide with those in which fracturing is most intense. The integration of fault/fracture modeling, well-production characteristics and the 9C3D survey results indicate the potential of the latter as a field-development tool in terms of optimizing well locations, well planning and reservoir management decision making.
In the northern part of the Sultanate of Oman, oil is produced chiefly from carbonate reservoirs of Cretaceous age. Major fields such as Fahud and Natih (Figure 1a) were discovered in the 1960s and rapidly brought on stream by Petroleum Development Oman (PDO). The oil-bearing reservoirs occur in the Natih and the Shu’aiba carbonate formations (Figure 2). The main reservoir in both fields is the Natih Formation. The Natih field was brought on stream in 1967 and, like the Fahud field (O’Neill, 1987), production was initially by a depletion drive. As a result of falling reservoir pressure, water injection was installed and this was later supplemented by gas injection (Dijkum and Walker, 1991). When the petroleum engineers of PDO recognized the importance of fractures to the development of these fields, the developments were changed to gas/oil gravity drainage (GOGD) (Bostock et al, 1990). The full-field GOGD involves lowering the fracture oil rim by 70 m through crestal gas injection and down-dip water production.
Essential to the success of the developments was the improvement of oil recovery from the fractured carbonate reservoirs. Core analysis and field studies had provided a semi-quantitative model of fracture frequency and it was recognized that the degree and orientation of fracturing obtained from 3-D seismic might prove an important tool. However, existing compressional seismic coverage could not give adequate information on fracture intensities and orientations and it was necessary to explore the use of a new seismic technique. As a result, it was decided to acquire a 9C3D (nine-component, three-dimensional) seismic survey in an appropriate candidate field. The objective was to examine whether the reservoir fracturing would manifest itself in shear-wave anisotropy, the measurement of which could then be used to characterize the direction and intensity of fracture systems.
The Natih field (Figure 1b) was selected as the candidate for the 9C3D trial seismic survey on the following grounds:
Technical—the existence of a fracture model, and of a single formation (the Fiqa Formation) between the target Natih reservoir and the surface.
Economic—a 3-D (compressional) survey was scheduled for the field.
In advance of the full survey, a small-scale 2-D pilot seismic survey (9C2D) was acquired in 1989 to test the feasibility of the full 9C3D. In addition, two vertical seismic profiles (VSPs) (a zero-offset and an offset VSP, both in Well N-85) were acquired and a wireline (dipole) shear log was run, both as independent measures of anisotropy (Hake et al, 1998). Positive results were obtained from these tests and the 9C3D survey trial was acquired in 1991.
GEOLOGICAL SETTING AND RESERVOIR CHARACTERISTICS
The Natih field was discovered in 1963 by the Natih-1 (N-1) well and was estimated to contain approximately 3,000 million barrels of stock-tank oil initially in place at 32° API, with an ultimate recovery of about 630 million barrels. The field is located within the Fahud Salt Basin (Figure 1a) on a domal structure measuring about 10 km by 6 km. It is bounded to the north by a reverse fault with a throw of about 1,000 m.
The main reservoir is the Natih Formation (Figure 2). It is a 380-m-thick carbonate that is subdivided into a series of stacked oil-bearing members, A to E. The bulk of the reserves, however, are held in the Natih-A and Natih-B. The cap rock is shale of the outcropping Fiqa Formation and the underlying strata are shales of the Nahr Umr Formation that cap the Shu’aiba reservoir (Figure 2).
Age and Lithology
The Natih Formation (Figure 2) is a middle Cretaceous (Albian to Cenomanian) sequence of chalky limestones of the Wasia Group (age equivalents of the Mauddud and Mishrif formations in the United Arab Emirates (Harris and Frost, 1984)). The formation consists of four main upward-shoaling cycles of argillaceous lime mudstones and wackestones coarsening upward into algal and/or rudist packstones. The Natih-A at the top of the reservoir is eroded and had been exposed and leached prior to the deposition of the overlying Fiqa Formation. Another subaerial exposure surface has been identified within the Natih Formation at the top of the Natih-E.
Summary Structural History
After the Natih was deposited, extension created NW-trending faults across the field. This was followed by a compressional phase that began in the Late Cretaceous and which re-activated the extensional faults as reverse faults. However, the main compressive event was associated with the uplift of the Oman Mountains during the Pliocene and there was also minor strike-slip movement along the northern segment of the Maradi Fault Zone (Figure 1a). This tectonic activity produced the NE-trending fault and fracture set (NE–SW compression) that dominates the area.
Reservoir quality is variable within the Natih Formation. The best reservoir conditions occur at the top of the shoaling-upward cycles in rudist-bearing rocks that have an average porosity of 25 percent, and which are partly leached (Davies and Niko, 1995). However, the permeability of the matrix is generally only in the range of 1 to 30 millidarcies (mD) and production from the matrix is almost entirely dependent upon the fracture network. Understanding the spatial distribution and orientation of the fractures is critical in optimizing the development of the reservoir. Fracture development largely results from the regional tectonics in combination with uplift, or from release of stress.
FAULTS AND FRACTURES IN THE NATIH FIELD
At the time of the 9C3D trial, much of the understanding of the various fracture mechanisms had been derived from surface geological observations. These observations were made on anticlines (for example, the Salakh Arch, about 50 km east of the Natih field) that are analogous to the Natih structure. Subsurface information was provided by core and borehole image logs of vertical wells in the Natih field. Fractures were seen to have a predominantly northeasterly trending strike direction, except in the east of the field where a northerly direction occurs in Well N-45 (Figure 3). The northeasterly trend was also seen in tracer tests where there was a strong preferential flow in that direction, although locally a minor northwesterly flow was also detected. Fracture attitude is consistently subvertical (Mercadier and Makel, 1991).
Since 1991, Formation Micro Imager (FMI) and Azimuthal Resistivity Imager (ARI) logging has been undertaken in horizontal wells. In addition, detailed fracture-modeling studies identified three types of extensional fractures:
NE-trending fractures produced by regional tectonics;
Folding or curvature-related fractures; and
Log evidence of fracturing due to local folding or horizon curvature was provided by data from the northern edge of the field. For example, FMI data from Well N-93 (Figure 3) show NW-trending fracture orientations, consistent with the local high curvature that is in excess of 2° per 100 m. Fault-related fractures seem to represent only a small proportion of the fractures at the surface and at reservoir depth. Two horizontal wells (N-112 and N-113; see Figure 3) both crossed minor normal faults with throws of less than 5 m. ARI data from Well N-112 indicated a cluster of nine fractures with an average spacing of 1 m in the footwall of one fault. The other well (N-113) showed a small, but insignificant increase in fracture intensity near another fault (Whyte, 1995). The most obvious contrast between the fracture system at reservoir depth (less than 700 m below ground level) and those described from outcrop studies, is the widespread development at the surface of an intense (generally NW-trending) fracture set. The general absence of these fractures in the subsurface is consistent with their relationship to uplift or stress release.
1997 Fault and Fracture Review
The 3-D survey of 1992 was interpreted in 1993 and Whyte (1994) published some of the field development findings. The map-set generated from this conventional 3-D survey was used in an integrated 1997 review of the fault/fracture system and production characteristics (Hitchings, 1997). The faults were identified as belonging to three distinct groups, as follows:
Main Fault Zone—the major field-bounding reverse fault and its associated faults (NW-orientation),
A set of ‘fishnet’ faults (NE- and NW-orientation),
Sets of en-echelon faults (generally NE-orientation) that may be bounded by faults with a curvilinear trace (in map view) or by reactivated NE-component fishnet faults.
Figure 4 is a schematic representation of the fault types that affect the Natih-A, -B, and -C units. The fault pattern identified from the interplay of the three types and the importance of strike-slip and block rotation has been evaluated. More importantly, a series of ‘domains’ was defined based on the fault characteristics and especially on the intensity and orientation of smaller faults within larger blocks (Figure 5).
A comparison of fracture orientations from cores and borehole images shows close agreement with the fault orientations (Figure 3). Blocks where rotation is interpreted have fracture strikes that depart from the northeasterly regional strike orientation by about 20°; for example, Well N-85. Fractures associated with the main fault and the steeper curvature are also evident (in Well N-93, for example).
The fault and fracture characteristics have a predictably close relationship with production behavior in that the most ‘disturbed’ blocks have the highest productivity. In terms of production, wells located within areas described as ‘quiet’ in terms of faulting, produce typically 535 barrels of oil per day (bopd) whereas a well in a ‘disturbed’ area can produce 3,780 bopd. Figure 6 shows an area of the field with the ‘productivity index’ (PI) (four classes) superimposed. The highest productivity (coded blue and purple) comes, for the most part, from a ‘rotated’ strike-slip block. The poorest productivity (coded green and pink) occurs in ‘quiet’ areas (in terms of the fault classification).
The 9C3D time-splitting areas delineated from the shear seismic survey have a clear correlation with the fault domains described above and will be discussed after the description of the 9C3D survey.
THE 9C3D SEISMIC SURVEY TRIAL
The purpose of the 9C3D trial was twofold:
Firstly, it had to prove the viability of shear seismic in its entirety; that is, acquisition, processing and interpretation for the detection and quantification of fractures.
Secondly, it had to satisfy a business need in terms of its impact on an oil development plan.
An area in the west of the Natih field was selected for the 9C3D survey (Figure 1b). This avoided the central area where existing production facilities may have hampered the survey quality, but it still represented a large part of the area of the field. The remainder of the field was to be covered by a conventional 3-D survey.
Potters et al. (1999), Hake et al. (1998), and Van der Kolk et al. (in press) have presented the results of the 9C3D survey. The 9C3D technique is considered by the geophysical community as one of the most promising innovations of recent years. However, what was missing was a sound understanding of the relationship between the observed seismic anisotropy on the one hand, and the Natih fracture pattern and production behavior on the other. In early 1997, a detailed review of the performance of all wells in the Natih field was conducted, and this has now been linked to a new and much more detailed model of faults and fractures. These new results stimulated a fresh look at the Natih 9C3D data, and our conclusions are the subject of this paper.
ANISOTROPY OF THE NATIH RESERVOIR
The principle of fracture detection by seismic shear waves is quite simple, and is related to optical birefringence as observed in crystals. A vertical shear wave polarized in an arbitrary direction impinging upon strata containing vertical fractures will split into a fast and a slow wave that can be thought of as propagating independently with different velocities (Figure 7). The wave polarized in the plane of the fractures is not affected by the presence of the fractures, but the one perpendicular to it is slowed down. After a while, both waves are reflected from the bottom of the fractured strata and travel upward, again with different velocities. More-intense fracturing produces a larger difference between the two velocities. When the signals are recorded at the surface, the time difference between the slow and the fast wave can be detected by unraveling the rather complicated mixed wave motion observed by horizontal particle-motion detectors (multicomponent geophones). For this purpose, special shear vibrators are used to generate waves in two perpendicular polarizations. This enables an accurate measurement of the direction of the fast and slow waves from a total of four wave fields, as each source direction produces a signal in two horizontal geophones.
All unwanted effects of shallower layers were removed by anisotropic stripping (Potters et al, 1999). As a result, an accurate interpretation of the Top Natih (top of the Natih reservoir) and Top Natih-E (an important regional seismic marker) was made and the anisotropic parameters for this interval were determined. Superimposing the slow and fast wave data, as in Figure 8, shows the anisotropy. In the Natih interval, the difference between the fast (SF) and the slow (SS) shear waves shows itself as the time (vertical) separation of the red and green reflections. Because of the time splitting across the two lines, it is possible to identify locations where the anisotropy is small and locations where it is large. Small time-splitting values (four milliseconds (ms)) are predominant on the left in Figure 8; that is, on the western and southern flanks of the field. A large increase in time splitting occurs up structure to the right, as shown by the growing time-distance between any given pair of red and green loops (locally greater than 30 ms). Moreover, close inspection reveals, both on the in-line (Figure 8a) and the cross-line (Figure 8b) sections, that the time splitting gradually increases with increasing depth starting from the Top Natih reflection. This is exactly what is expected from a fractured interval as deeper reflections have traversed a larger interval over which the slow wave gets retarded relative to the fast wave.
Taking the 200-ms interval from Top Natih to Top Natih-E (these being the strongest and most reliable horizons) as representative of the reservoir, the anisotropy direction and time splitting over the entire survey area can be determined. The time splitting is divided by the two-way interval time for the fast wave to obtain relative time splitting. Figures 9a and 9b are map views of the direction of the fast wave in the Natih reservoir and the percentage time splitting for the same interval, respectively.
The anisotropy of the Natih interval is characterized by a polarization direction for the fast horizontal shear wave that trends approximately northeast, as shown in red on Figure 9a. As this direction coincides with that of the pervasive open fractures, it is in agreement with geological expectations. There are also areas where the fast-mode polarization tends more to the north-northeast (60° – 80°) and east-northeast (25° – 40°) and where the fractures are probably oriented somewhat differently.
With reference to Figure 9b, there is a large shear-wave anisotropy present over about half the survey area. The anisotropy is large both in absolute (locally greater than 30 ms) and in relative terms (more than 15%). Reliable high-splitting values do not coincide with mapped faults, which excludes fault-related extensional fractures as a cause of the anisotropy. However, faults do bound several domains of different shear-wave splitting within which the splitting variations are relatively small. This could imply that each domain has its own fairly uniform fracture pattern and intensity, whilst being in a different mechanical stress state from its neighbors.
LINKING SEISMIC ANISOTROPY AND GEOLOGY
Anisotropic areas can be identified by comparing the time-splitting map (Figure 10b) with the modified domain map (Figure 10a) of the 9C3D survey. Areas having the highest degree of time splitting or anisotropy are seen to coincide with ‘rotational’ areas, or those that are interpreted as having been subjected to strike-slip movements. These are, in turn, the areas of highest production and the most fractured areas. In contrast, the areas with the lowest degree of splitting coincide with ‘quiet’ areas on the domain maps. The limitation of areas of different time splitting (anisotropy) by faults is then explained as the bounding of domains that have differing degrees and orientation of faults and fractures.
ECONOMIC APPLICATION OF 9C3D
Given the interdependency of seismic anisotropy, structural domains and production behavior, it is feasible to suggest that the 9C3D map could be employed as a planning tool to locate potential high-productivity wells. Mueller (1992) described how shear waves were used to predict the lateral variability in vertical fracture intensity in the Austin Chalk of Texas. In this case, comparing interpreted stacked SF (fast) and SS (slow) sections enabled anomalies to be seen in the shear-wave data. Anomalous areas, where there are lateral seismic discontinuities in the SS sections, were interpreted as fracture zones about 250 m wide. Such a zone was tested with a horizontal well drilled through two SS amplitude anomalies. The results showed two fracture zones along the well trajectory that coincided with the SS zones. In the same way, the Natih 9C3D dataset could be used as a well planning tool by:
Selecting a general area for locating a well based on production data.
Planning and refining the well track using SF and SS sections generated as random lines from the nine-component data cube in order to optimize the intersection of fractures.
It should also be noted that domains of small shear-wave anisotropy might be areas where non-GOGD production is possible and that the optimal development may be by localized water flood.
The 9C3D survey results are consistent with the 1997 fault/fracture model but they cover only the western part of the field. Whyte (1995) described developing the movement of the fracture oil rim as the challenge of ‘hitting a moving target’. As the oil rim is further developed and lowered, the well-targeting zone will move down the flank of the field. A careful consideration of the business case may show that completing the 9C3D survey over the whole field could have a cost benefit for mitigating the risk of siting wells in suboptimal areas.
The 3-D (compressional) seismic survey has been used to define areas of differing fault and fracture characteristics, here termed ‘domains’.
Areas with variable shear-wave splitting in the 9C3D seismic survey can be related to different domains.
As the domains show distinctly different production characteristics in terms of well productivity, there is the potential for using the 9C3D as a tool for measuring the intensity of fracturing and hence as a well location tool.
A review of the business case may show that completing the 9C3D survey over the whole field could have a cost benefit for lowering the risk of siting wells in suboptimal areas.
The authors would like to thank Petroleum Development Oman and the Ministry of Oil and Gas of Oman for permission to publish this paper. VH would like to thank the members of Petroleum Development Oman’s Fahud/Natih Petroleum Engineering Team for their support in the development of the fault/fracture model. Thanks are also due to two anonymous reviewers whose suggestions were appreciated. Pascal Richard is gratefully acknowledged for his encouragement and enthusiasm.
ABOUT THE AUTHORS
Victor (Vic) Hitchings was awarded a PhD in Geology by University College, Swansea, U.K. in 1982. He joined Shell as a Reservoir Geologist in 1986 after six years in core evaluation with Robertson Research in the U.K. and the Far East. After working in the Shell International Exploration and Development Research Laboratory in The Netherlands and for Shell Nigeria as a Reservoir Geologist, he joined Petroleum Development Oman in 1993. Vic was Senior Production Geologist in the Fahud/Natih Team from 1993 to 1997, working primarily on fractured reservoir development by Gas/Oil Gravity Drainage, and was the Development Geologist for the Natih Team from 1996 to 1997. He is at present a Senior Reservoir Geologist at Shell Technical, Exploration and Production, in Rijswijk, The Netherlands.
Hans Potters joined Shell in 1984 after obtaining his PhD in Physics from Utrecht University. He has held geophysical research and management positions in The Netherlands, U.S.A. and Oman. From 1991 to 1993 he was responsible for the Natih 9C3D project carried out by Petroleum Development Oman (PDO) and Shell Research. In 1994, he was appointed head of the Quantitative Interpretation Group of PDO. Hans is now the Subsurface Integration Coordinator at Shell Technical, Exploration and Production in Rijswijk, The Netherlands.