The Khuff-C reservoir in the Ghawar field is a stratified sequence of cyclic carbonate-evaporite deposits within the Permian Khuff Formation. The reservoir is heterogeneous, complex, and influenced by syndepositional diagenesis. Wells drilled into the Khuff-C in the ‘Uthmaniyah sector of Ghawar are usually prolific producers of non-associated gas but some have intersected poor-quality reservoir intervals with little or no gas production. The Khuff-C reservoir rocks were deposited in a peritidal setting where slight changes in sea level created locally exposed highs. The exposure in an arid climate resulted in outliers of porosity occlusion formed by evaporite cements within the Khuff-C reservoir. The outliers are variably sized and randomly distributed and the challenge is to predict their occurrence in order to avoid them in development drilling. Inverse modeling of the ‘Uthmaniyah 3-D seismic data has identified the tight-porosity outliers as areas of anomalously high acoustic impedance. Integration of 3-D seismic analyses with petrophysical and other well data has improved the reservoir characterization and reduced the drilling risk.
The supergiant Ghawar field is located in eastern Saudi Arabia in the eastern part of the Arabian Platform (Figure 1). The Permian Khuff-C carbonate gas reservoir in the ‘Uthmaniyah sector of the Ghawar field was discovered in 1982. The reservoir is a prolific producer of non-associated gas and some condensate. In addition to the Khuff-C, the Khuff-A and Khuff-B reservoirs also produce gas in the Ghawar field. The Lower Silurian Qusaibah shale is considered to be the source rock for the Khuff reservoir. A mid- to late-Mesozoic phase of gas yield charged the Khuff traps that existed prior to the formation of the Zagros uplift (Bishop, 1995). The hydrocarbon migration pathways to the Khuff reservoirs are believed to be deep-seated vertical faults as seen in 3-D seismic data.
The reservoir porosity and permeability distribution is complex and is primarily controlled by the distribution of the depositional facies and by later diagenesis. The Khuff sediments were deposited in a subtidal and gently dipping tidal flats in an extremely arid environment. The Khuff-C reservoir facies vary in their reservoir quality as a result of their diagenetic history. Khuff-C porosity varies from about 30 percent to less than 5 percent, which is the economic cutoff. Parts of the reservoir facies have been dolomitized and later leached, and this resulted in good reservoir development. Some dolograinstone facies, however, have been cemented by anhydrite that destroyed the reservoir porosity and permeability.
These variably sized anhydrite-cemented ‘outliers’ within the Khuff-C reservoir facies are distributed randomly throughout the field. The challenge is to predict their occurrence between wells in order to avoid them during the development-drilling program.
Khuff-C reservoir well spacing in this area of Ghawar is at least 2 km. With such a coarse well spacing it is difficult to characterize the reservoir heterogeneity from well data alone. The integration of 3-D seismic data with petrophysical information and the reservoir simulation history has improved our understanding of the reservoir complexity and the porosity distribution of the Khuff-C reservoir. This has permitted us to define the local tight porosity facies of the Khuff-C reservoir throughout the field. Amplitude inversion of the 3-D seismic data corresponded to high acoustic impedance (velocity x density) in the vicinity of wells that intersect tight Khuff-C reservoir intervals. The seismic inversion results also indicated the localized high acoustic impedance-low porosity anomalies (or outliers) that are distributed sporadically throughout the field. These results were used to locate development wells over areas with good reservoir quality and to avoid tight facies. Subsequent drilling results have been extremely successful––all 10 wells that have been drilled using the integrated reservoir model, intersected the porous and productive Khuff-C reservoir as predicted (Dasgupta et al., 2001). The results also verified the interwell heterogeneity of the reservoir properties predicted from seismic inversion (Dasgupta et al., 2000).
The Khuff Formation represents the earliest period of major carbonate sedimentation in the area of the Ghawar field, resulting from the Permian transgression of the Neo-Tethys Ocean over the eastern part of the Arabian Plate (H.S. Talu and F.A. Abu Ghabin, 1987, unpublished Saudi Aramco Report no. 117). The generalized stratigraphic column (Figure 2) shows the stratigraphic position of the Khuff Formation. During the Permian, the Arabian Plate was part of Gondwana near latitude 30°S (Bambach et al., 1980) and most of the Arabian Peninsula was covered by a restricted evaporite-carbonate shelf-platform sea.
Five depositional members were recognized within the Khuff Formation in Saudi Arabia (Al-Jallal, 1995); in descending order they are, Khuff-A, -B, -C, -D, and -E, as shown in the stratigraphic column Figure 2. The Khuff-A to -D members show cyclic sedimentation. Each cycle starts with a gradual transgression of subtidal grainstones that make up the Khuff reservoirs and ends with the gradual deposition of a regressive unit of intertidal and supratidal muddy and evaporitic deposits that constitutes the seal. In contrast, the transgressive Khuff-E consists of the ‘basal Khuff clastics’.
The Khuff Formation is within Arabian Plate tectonostratigraphic megasequence AP6 of Sharland et al. (2001), the base of which is the pre-Khuff unconformity dated at 255 Ma. The depositional cycles in the Khuff Formation can be correlated with the global Late Permian eustatic cycles of Haq et al. (1987). Based on the correlation of maximum flooding surfaces, the Khuff-A, -B, -C, and -D cycles were identified by Sharland et al. (2001). The Khuff-C reservoir is within the Khuff-C member in a sequence of cyclic carbonate-evaporite deposits. The reservoir interval consists of interbedded, tight and porous limestone and dolomite beds sandwiched between thick anhydrite-rich layers that act as seals.
Depositional Environments and Facies
Figure 3 shows the major facies and environments of deposition for the Khuff Formation (Al-Jallal, 1995). The depositional environments were predominantly intertidal and sabkha together with subtidal carbonate sand shoals, lagoons, and sandbars. The arid conditions that exist today along the southern shore of the Arabian Gulf prevailed at the time of deposition of the Khuff Formation. Sabkha features similar to those of today have been seen in Khuff cores (Al-Jallal, 1989).
Several wells in the ‘Uthmaniyah sector of the Ghawar field have penetrated the prolific Khuff-C gas reservoir (Figure 4). Two of these wells, A and B, intersected tight Khuff-C reservoir intervals. Cores from the Khuff-C in these wells show that anhydrite has cemented the dolograinstone matrix. To the east of wells A and B, a stratigraphic zero porosity edge was interpreted as the limit of preservation of Khuff-C reservoir porosity in the area. A west-to-east cross-section through the field, based on logs from wells 16, 1, and B, illustrates the heterogeneity of the reservoir (Figure 5) with tight well-B being on its eastern margin. Based on this original interpretation of the well control, the eastern flank of the Khuff anticline in the area was considered to be tight (Figure 4) and was removed from developmentdrilling plans for the Khuff-C reservoir. However, this interpretation has recently been revised through the integration of 3-D seismic and petrophysical data described in this present study.
The cores of the Khuff-C reservoir interval from several wells in the ‘Uthmaniyah area were studied for composition, texture, lithology, facies, porosity type, and cementation, and were calibrated to wireline logs. Thin sections from the cores were examined for textural details and for diagenetic overprints. Cores from porous Khuff-C reservoir intervals are composed of grainstone, dolograinstone, and dolomudstone (Figure 6). Lateral depositional facies regionally control the reservoir distribution. Localized diagenetic changes played a dominant role in the porosity distribution as cementation has, in places, destroyed the reservoir porosity (Talu and Abu-Ghabin, 1989). Microscopic examination of thin sections from Khuff-C reservoir cores show intergranular and intercrystalline porosity in well-3 that contrasts with core from well-B in which anhydrite cement occludes the intergranular pores (Figure 7). Anhydrite cementation is a common phenomenon in the Khuff reservoirs. It usually occurred almost contemporaneously with deposition as saturated fluids percolated downward and precipitated anhydrite in the underlying grainstone pore spaces and in the intercrystalline porosity of mudstones. The grainstones of this type are usually dolomitized. A simplified model of the diagenesis and cementation process is shown as Figure 8.
INTEGRATION OF RESERVOIR CHARACTERISTICS
Sonic and density logs from existing wells in the area were processed for petrophysical information and field-wide calibration and correction. Stratigraphic markers at and near the objective reservoir zone were correlated on wireline logs for the wells and converted to two-way travel time to aid the interpretation of seismic data.
3-D seismic data were acquired in order to understand the reservoir framework in the study area, delineate the complexity of the reservoir, and map the porosity distribution. The seismic survey parameters were designed to better image the deeper Khuff and pre-Khuff objectives down to 16,000 ft. Five vibrators were used at each source point, with a 12-sec up-sweep and a sweep frequency of 8 to 80 Hz. The cable length was 3,600 m. The field geometry created a 144-fold stack with 24-fold in-line and 6-fold cross-line. The surface coverage was 200 source points per sq km and the Common Depth Point bin size was 25 by 25 m. The total seismic survey area was approximately 850 sq km. The data were processed through a stratigraphic flow sequence in order to preserve the relative amplitudes. The final Dip Moveout-processed, time-migrated seismic stack volume was loaded onto workstations for interactive interpretation.
The seismic to well-log calibration was a critical step in the amplitude inversion process that included wavelet extraction and the generation of synthetic seismograms to tie recorded seismic traces near each well. This calibration was performed for all wells in the area that had sonic and density logs.
Wavelets were extracted using the available well-log interval and the impedance logs and stack seismic traces over a window centered on the reservoir. The wavelets derived from the well-log calibration indicated a consistent waveform similar to a Ricker wavelet of 35 Hz. The phase variation for the wells analyzed was within 20° of the theoretical Ricker wavelet. Minor variations in the extracted wavelets from well to well were believed to have been caused by residual random and coherent noise in the seismic data. A representative wavelet from well-3 was computed for the seismic volume, and synthetic seismograms and acoustic impedance were computed along the target reservoir level for the entire seismic volume. The wells were calibrated to tie the picked geological markers to seismic horizons and a wavelet was derived from the seismic-to-well calibration.
The Khuff-C reservoir is 170 to 250 ft thick in the depth range of 12,000 to 14,000 ft in the study area. This is within the resolution of the 30 Hz dominant frequency of the seismic data. The interpretation of the seismic data showed an abrupt termination of amplitude in the vicinity of wells A and B that had tight reservoir intervals.
Synthetic seismograms generated from the well logs indicated a sharp waveform contrast and an amplitude trough associated with wells having porous Khuff-C. Tight Khuff-C intervals corresponded to high acoustic impedance and diminishing amplitude or, in some cases, to a reversal of polarity. The sharp waveform contrast that were observed in the synthetic seismograms from porous to tight Khuff-C reservoir intervals (Figure 9), provided a basis for delineating porosity variations within the reservoir.
Figure 10 shows seismic line X–X’ through well-3 (porous Khuff-C) and well-B (tight Khuff-C) approximately 2.5 km to the east (Figure 4). The seismic data show strong amplitudes that correspond to the Khuff-C reflector in well-3. They terminate abruptly to the east of well-3 but reappear to the east of well-B. The acoustic impedance is low for the seismic line at well-3 and increases abruptly to the east in the vicinity of well-B. This suggests that amplitude diminishment is localized in the vicinity of wells that have tight Khuff-C porosity.
An initial model-based 3-D seismic amplitude inversion was performed on the data using calibrated sonic and density logs that were available in 17 wells. The initial model assigned an impedance log for every trace in the seismic volume by interpolating impedance values between the well-control points. The iterative inversion process to match the amplitudes for each trace in the seismic-volume time window then updated the model. The amplitude inversion was thus target oriented, iterative, and constrained by acoustic impedance computed at the wells.
Acoustic impedance and reservoir porosity
Areal distribution of the Khuff-C impedance (Figure 11) illustrates the heterogeneity of the reservoir quality; this information would not have been available from well data alone. The Khuff-C reservoir porosity was predicted from the 3-D inversion with blue and green representing higher levels of porosity, and yellow and red showing low levels. The acoustic impedance versus porosity cross plots from well logs indicate a linear trend. An example of a tight location is well-16 in Figure 12, this being one of the wells drilled before the seismic inversion modeling. The computed impedance indicates a highly variable reservoir quality along the line. A porous location was identified in the continuous low-acoustic impedance zone shown in Figure 13. The corresponding seismic line and the acoustic impedance for the initial model derived from the wells and interpolated between wells are also shown.
The acoustic impedance map (Figure 11) shows that high impedance/low porosity is localized in the vicinity of the two tight wells A and B. The area to the east of these wells, originally interpreted to be beyond the limits of Khuff-C reservoir porosity, shows good porosity (greens and yellows) on the impedance map. These results imply that Khuff-C porosity extends to the entire eastern flank of the ‘Uthmaniyah sector and thus opens up a large fairway for porous Khuff-C reservoir delineation (Dasgupta et al., 1999, 2000). Thus, the Khuff-C gas reservoir volume in this area is significantly larger than was originally interpreted based on the well data alone.
Observed reservoir pressure versus simulation
The original reservoir simulation of the Khuff-C gas reservoir in the study area used a geological model with zero porosity thickness to the east of wells A and B that intersected a tight Khuff-C reservoir interval. Figure 14 shows the reservoir simulation model study for the period from the start of gas production in 1984 until 1998. The results indicate that the observed reservoir pressures (ranging from 7,000 to 6,200 psi) were consistently higher by 700 to 800 psi than the modeled pressures (green curve in Figure 9). This additional pressure suggests extra sources of Khuff-C reservoir energy (P. Hsueh and M. Al-Shammari, 1996, unpublished Saudi Aramco Report RSD3.058). In order to achieve a history match with the observed pressure at the wells, the gas saturated porosity thickness ϕ – h × Sg (where ϕ = porosity; h = thickness; and Sg = gas saturation) of the reservoir was globally increased by 40 percent. The modeled pressure profile (blue curve in Figure 14) at the new reservoir volume can be explained by increasing the pore volume or reserves. The inclusion of reservoir porosity beyond the original zero porosity accounts for increased pore volume of the Khuff-C reservoir.
Development Drilling Program
The distribution of impedance of the reservoir interval was used as a guide for predicting reservoir quality and in assisting future well placement. The acoustic impedance map (Figure 11) was used qualitatively as a guide in planning the sub-surface location of development drilling program. The results of the development drilling were highly successful. Ten wells have been drilled so far using the acoustic impedance from seismic inversion and each of them intersected porous Khuff-C gas reservoir intervals (Figure 15).
Khuff-C reservoir heterogeneity and localized tight facies were verified in a recent well in the ‘Uthmaniyah area. Evaluation well-18 was located before the integration of seismic data (see Figure 4). Sonic and density logs from the well had showed virtually no porosity in the Khuff-C (Figure 16). However, the seismic data showed a local amplitude anomaly at well-18 that was surrounded by normal amplitudes and low acoustic impedance corresponding to porous Khuff-C. This suggested that the well had intersected a tight Khuff-C ‘island’ (see Figure 8) but was surrounded by good quality reservoir. Based on the interpretation of the 3-D seismic data, well-18 was tested over the Khuff-C reservoir interval and yielded a flow rate of nearly 15 million cu ft of gas per day (Figure 16). Fractures in the reservoir that intersected the wellbore permitted gas to flow inspite of the tight Khuff-C in the well. The well, that would previously have been plugged and abandoned, will now be put into production.
The integration of 3-D seismic acoustic impedance, petrophysical data from wireline logs, core analysis, geological interpretation, and reservoir simulation history-match results, has redefined the stratigraphic porosity edge of the Khuff-C reservoir over the ‘Uthmaniyah sector of the Ghawar field. The integration of observations from various disciplines has resulted in a synergy that has reduced, if not eliminated, the ambiguity in the interpretation of results from each individual discipline. The stratigraphic porosity edge, interpreted from the well data on the eastern flank of the field, has been revised. The results conclude that Khuff-C porosity extends to the entire eastern flank of Khuff anticline. This improved reservoir characterization model has increased the pore volume of the Khuff-C reservoir in the area and has added significantly to the gas and condensate reserves. Reservoir pore volumes in the new model also provide a more accurate prediction of reservoir performance. Since the completion of this study, 10 wells (K-1 to K-10 in Figure 16), have been drilled in the area with the Khuff-C gas reservoir as the objective. Each well location was optimized to penetrate the low impedance/high-quality Khuff-C reservoir. All 10 wells have encountered porous reservoir intervals as predicted by the model.
We thank the management of Saudi Aramco for their support and for permission to publish this paper. George Grover, Mohammed Amoudi, Abdul-Jaleel Al-Khalifa and the members of the Khuff Gas Evaluation team are thanked for their suggestions and comments during the study. We also thank GeoArabia’s editors and two anonymous reviewers who greatly improved the paper. The design and drafting of the final figures was by Gulf PetroLink.
An unrefereed version of this paper (Reservoir characterization of Permian Khuff-C carbonate in the supergiant Ghawar Field of Saudi Arabia, by Shiv N. Dasgupta, Ming Ren Hong and Ibrahim A. Al-Jallal) was published in The Leading Edge, v. 20, no. 7, p. 706–717 (July 2001).
ABOUT THE AUTHORS
Shiv N. Dasgupta is a Geophysical Consultant in the Reservoir Characterization Department of Saudi Aramco. He has advanced degrees from St. Louis University, Washington University, St. Louis and Southern Illinois University. He has been employed with Saudi Aramco since the early 1980s. He previously worked in exploration for Amoco Production, Mitchell Energy & Development, and Conoco. Shiv’s professional interests include the integration of geophysical techniques for reservoir development and production optimization.
Corresponding author. E-mail DASGUPSN@aramco.com.sa
Ming-Ren Hong is a Geophysical Consultant with Saudi Aramco. He has a BSc in Atmospheric Physics (1973) and an MSc in Geophysics (1977) from the National Central University, Taiwan. He received his PhD in Geophysics from the University of Texas at Dallas in 1982. He joined Saudi Aramco in 1991. He had worked at the Center for Lithospheric Studies, University of Texas at Dallas as a Research Engineer from 1982 to 1984, and from 1984 to 1991, he was a Senior Research Geophysicist with Arco Oil and Gas Company. Ming-Ren is a member of SEG and SPE. His professional interests include seismic modeling and inversion, reservoir characterization, and integrated interpretation.
Ibrahim A. Al-Jallal is Chief Geologist of Southern Area Reservoir Characterization in Saudi Aramco. He received his BSc in Geology/Chemistry from King Saud University, Riyadh (1973) and an MSc in Geology from Western Michigan University (1979). He was awarded a PhD in Petroleum Geology from Imperial College, London in 1990 for a study of the Khuff Formation. His PhD study was the deposition, diagenesis, and reservoir prediction of the Khuff Formation in Ghawar field, Saudi Arabia. He recently extended this study regionally to include all the Gulf States. Ibrahim has experience in wellsite operation, reservoir development, prediction, layering and depositional modeling. His involvement in geological R&D and reservoir characterization projects have included reservoir characterization support for field development and geological studies, and participation with international consortia in E&P activities. He was in charge of a team that studied the Jauf reservoir in the North Ghawar Reservoir Characterization project that included implementation of sequence statigraphy and integration of 3-D seismic interpretation. In 2000, he became the Chief Geologist for Southern Area Fields Characterization that includes Ghawar, Abqaiq and the Arab-D in central Saudi Arabia.