P-wave attenuation helps identify lithology and pore-fluid type - Statistical Data Included

A 3-D field example illustrates how P-wave attenuation was used to characterize an oil-prospective, turbiditic-sandstone reservoir in the Eastern Venezuelan basin

In this 3-D field example, P-wave attenuation was used to detect pore fluid (oil) in a deep, oil-prospective, turbiditic-sandstone reservoir in complex geology. Amplitude and attenuation data are compared and the results shown. Petrophysical data from a lone well were combined with seismic attributes to generate maps that showed attenuation to be a better indicator of pore fluid than amplitude. In addition, the attenuation method revealed new potential prospects.

OBJECTIVE

Several studies show that seismic attenuation is a promising seismic attribute for distinguishing fluid type in sandstone reservoirs,[1,2] monitoring steam-heated reservoirs,[3] detecting gas in channel sands,[4] discriminating lithology as a function of saturation[5] and estimating permeability from cores using ultrasonic seismic data.[6,7]

Based on these experiences, an important play in the Eastern Venezuelan basin was selected as a pilot project to validate using P-wave seismic attenuation to detect pore fluid (oil). The aim was to characterize a deep, Miocene, oil-prospective, turbiditic-sandstone reservoir. Through RMS (Root Mean Square)-attenuation maps and attenuation-petrophysical relationships, the project sought additional information about this reservoir and identification of other prospects. This was a challenge because of structural and stratigraphic complexity and data from only one well (W-5) in this area.

After comparing amplitude and attenuation data from the target and other sandstone intervals at Well W-5, P-wave attenuation was shown to be more sensitive to pore fluid (oil) than P-wave seismic amplitude. Additionally, the RMS-attenuation map showed a better interpretative image of the target reservoir and provided other potential prospects in the study area. It suggested that seismic attenuation could be a better attribute for identifying oil prospective, turbiditic-sandstone reservoirs than conventional seismic amplitude in this basin.

FIELD BACKGROUND

The study area is located in the Eastern Venezuela basin in Northern Monagas, Venezuela, Fig. 1. Before reaching its primary target, an Oligocene sandstone, Well W-5 encountered a productive interval of Miocene, turbiditic sandstone between 15,901-15,950 ft. The sandstone contains 41.9 [degrees] API oil and is overlaid by a massive, 3,700-ft-thick shale that acts as a regional seal. Deposition of this turbidite body was controlled by currents and sea-bottom topography. At Well W5, both agents generated an elongated, ponded, turbidite body that was deformed by faulting and folding during the Miocene.

[Figure 1 ILLUSTRATION OMITTED]

After this unexpected discovery, a 3-D, P-wave survey was acquired to improve prior structural and stratigraphic interpretations based on 2-D seismic data. Several seismic attributes were tested to characterize the target reservoir. From the results, RMS amplitude was initially selected as the most sensitive attribute to the presence of turbiditic-sandstone bodies.

Although the RMS-amplitude maps showed a lithologic image--where higher amplitudes are associated with turbiditic sands and lower amplitudes with shales--it did not show a conclusive image of the target. However, from these maps, some general amplitude trends were interpreted as turbiditic channel, ponded and fan depositional facies. Additionally, the target reservoir's amplitude behavior at Well W-5 was a lithologic response not directly related to the presence of fluid (oil).

Based on increasing numbers of papers reporting variations of seismic attenuation with lithology, pore fluid and permeability at different frequency scales, P-wave attenuation was chosen as an alternative attribute to improve images and characterize the reservoir at Well W-5.

METHODOLOGY

To obtain a P-wave-attenuation trace from post-stack data, a modified version of the algorithm published by Mitchell[4] was applied on a 3-D, P-wave volume acquired in Northern Monagas. This technique is based on the power spectral ratio method (currently used in labs to obtain quality-factor values for cores). The spectral amplitudes (in Fourier domain) from two different time-windowed reflections are divided and plotted as a function of frequency, Fig. 2. This plot reveals a linear trend, the slope of which is directly related to the attenuation coefficient, or quality factor, Q.

[Figure 2 ILLUSTRATION OMITTED]

The present approach introduces a small change in this algorithm by considering background and interval spectrums. The background spectrum is calculated for a given trace between datum-to-maximum recorded times; thereafter, it is used as a constant value for each time sample along that trace. In contrast, the interval spectrum is obtained from a time-windowed reflection, the limits of which are controlled by the phase-skip criterion.[8]

In this way, the resultant time windows will present different widths, varying in time, depending on where the time skips occur in the instantaneous-frequency trace. This procedure avoids signal interference that could bias the estimated attenuation values, which are sometimes present after applying a moving, single-width, timewindow scheme. Additionally, the window-width selection is automatic and not user dependent. The linear least-squares method is applied on the back ground/interval spectral ratio vs. frequency for this time window, providing an interval-attenuation value. This process is repeated for all time windows along a given trace to obtain a blocky attenuation trace.

ATTENUATION VS. AMPLITUDE

Attenuation traces were inverted from a 3-D, P-wave-migrated volume for testing the attenuation response to the presence of lithology and pore fluid. Well W-5 is located at an anticline fold and limited by a system of reverse faults with a strike of about N70 [degrees] E, Fig. 1. The well was used in this study because it penetrates the oil-saturated sandstone reservoir between 15,905-15,950 ft (3.70-3.72 s) and other non-productive, turbiditic sand stone and shale intervals at varying depths. The gamma-ray log shows a sharp contrast between shale and sandstone, which provides a good seismic response associated with the turbiditic sandstones, even though these sandstones are under tuning (about 45 m).

The P-wave, inverted-attenuation section for line 425, which contains Well W-5 at trace 805 (Fig. 3), shows that the light-oil, turbiditic-sandstone reservoir is characterized by a significant attenuation anomaly compared with other sandstone/shale intervals along this well. The lateral extent of this anomaly, estimated at 330 ft, is stronger on the attenuation map (Fig. 3b) than it is on the amplitude map (Fig. 3a) for this oil-prospective sandstone reservoir (3.70-3.72 s).

[Figure 3 ILLUSTRATION OMITTED]

A detailed view of seismic amplitude and attenuation surrounding the target at Well W-5 is shown in Fig. 4. Again, it is evident that an anomalous attenuation value is located at the target, while the non-productive sandstone intervals have a flat amplitude behavior. However, the envelope amplitude shows flat behavior between 3.65-3.825 s in Fig. 4, which, in the RMS sense, is higher than is associated with the sand/shale intervals between 3.50-3.60 s. The conclusion from this comparison is that attenuation is a very sensitive attribute to the presence of oil in this turbiditic reservoir, but amplitude is not.

[Figure 4 ILLUSTRATION OMITTED]

A SYNTHETIC TEST

Since the oil-prospective reservoir at well W-5 is under tuning, P-wave amplitude and attenuation data could be a tuning-related effect rather than a lithology/pore-fluid effect. In fact, when a layer is below resolution, reflections from top and base of the layer are affected by destructive interference; the resultant reflection layer shows low amplitude, while its high frequencies are significantly attenuated.

For testing this effect, a synthetic seismogram was generated by using both density and P-wave sonic logs from W-5. These were combined with a zero-phase ticker wavelet with a central frequency of 17 Hz (close to that observed from seismic data). This seismogram was correlated and inverted to attenuation values for comparison with the real attenuation trace at Well W-5. If the attenuation anomaly associated with the target in line 425 were related to a tuning effect, it should be present in the synthetic-attenuation trace.

The comparison between real- and synthetic-attenuation traces shows that an attenuation increment appears at target level in the synthetic-attenuation trace. This is a result of the tuning effect and lithology, Fig. 5, (sandstone has high attenuation, while shales have low values). However, its appearance is different from the anomaly observed on real trace 805. From this result, it was concluded that an additional parameter, different from tuning, is causing P-wave attenuation in the oil-prospective reservoir. This additional parameter could be related to the pore fluid (oil).

[Figure 5 ILLUSTRATION OMITTED]

TURBIDITE-BODY CHARACTERIZATION

RMS amplitude and attenuation maps were generated from the 3-D, P-wave volume to compare images of the oil-prospective, turbiditic reservoir around Well W-5, Fig. 6. These maps represent RMS values calculated from top to base of this reservoir (0.02 s thick), following the time-structural seismic interpretation. The RMS-attenuation map shows pronounced improvement in the reservoir image compared to the RMS-amplitude map.

[Figure 6 ILLUSTRATION OMITTED]

Additionally, it is evident from the RMS-attenuation map that the attenuation anomaly at W-5 could be interpreted as an elongated sandstone body in a channel. Another elongated body is observed to the north, the attenuation anomaly of which is similar to the one obtained at W-5. It could indicate another oil-prospective, turbiditic reservoir. In contrast, the RMS-amplitude map shows both bodies isolated, with different average-amplitude values.

For characterizing this turbiditic-sandstone channel, various crossplots--including amplitude, attenuation and petrophysical data--were used to estimate net sand, porosity, clay volume and permeability, Fig. 7. All of these petrophysical properties represent average values calculated for each sandstone interval at Well W-5, except for net sand, which represents cumulative net sand from each sandstone subinterval. Here, because of limited, vertical seismic' resolution, some time windows contain several sandstone intervals; thus, it was necessary to sum the net sand thickness of each interval to represent overall net sand. The target reservoir has the following average properties: net sand = 45 ft, [Phi] = 11.7%, [V.sub.sh] = 18.0%, k= 25 mD and [S.sub.w] =33.5%.

[Figure 7 ILLUSTRATION OMITTED]

Analysis of these crossplots indicates that attenuation increases as cumulative net sand increases. In contrast, amplitude does not show any correlation with net sand, Fig. 7a. Correlations of amplitude and attenuation with porosity and clay volume are low, with regression factors between 0.24 and 0.0001, Figs. 7b and 7c. However, the attenuation-permeability crossplot shows a moderate correlation ([R.sup.2] = 0.47), Fig. 7d, indicating that attenuation increases with permeability. A similar result was obtained in laboratory tests by measuring attenuation from sandstone core.[6,7] In contrast, the amplitude-permeability relationship shows a very low correlation of [R.sup.2] = 0.17, Fig. 7d.

Because P-wave attenuation showed a reasonable correlation with net sand and permeability, both petrophysical properties were combined to create the fluidity-rock-quality term; this is a measure of oil mobility in sandstone. This term is the product of net sand and permeability.

A logarithmic curve was obtained from the attenuation fluidity-rock-quality crossplot, Fig. 8a, with a moderate regression of [R.sup.2] = 0.48. This regression curve was used to obtain the RMS fluidity-rock-quality map, Fig. 8b, from the RMS-attenuation map in Fig. 6. Although this fluidity map does not provide additional information about reservoir geometry, it shows three other facts: 1) a high probability that the reservoir contains oil, which is due to its high permeability rather than net sand thickness; 2) the anomaly associated with the target suggests an elongated body of smaller lateral extent than that observed on the P-wave attenuation map in Fig. 6b; and 3) it shows other potential oil reservoirs to the north of Well W-5 based on a high-fluidity response.

[Figure 8 ILLUSTRATION OMITTED]

CONCLUSIONS

It has been shown that, in the Eastern Venezuelan basin, P-wave attenuation is a more reliable seismic attribute than amplitude for detecting pore fluid in an oil-prospective, turbiditic-sandstone reservoir. Additionally, P-wave attenuation provided a more coherent image of this reservoir, improving its interpretation.

Although this conclusion is based on a single well, the correlation between P-wave attenuation with net sand and permeability suggests that, through fluidity-rock-quality maps, attenuation could help identify other potential reservoirs in the study area. Calibration of this kind of map with future well control should indicate prospective areas with significant oil mobility.

ACKNOWLEDGMENT

Thanks to PDVSA E&P for its support and permission to publish this article; Eduardo Alvarez and Lucas Zamora, from PDVSA E&P, for their enlightening conversations; and special recognition to Ian Gath, from PDVSA E&P, for valuable help with software and data manipulation. This article is an extension of SEG paper 1423, presented at the SEG International Exposition & Sixty-Eighth Annual Meeting in New Orleans, Louisiana, 1998, and an adaptation from the oral presentation, "Latin American Exploration Plays and Development," AAPG Annual Meeting in San Antonio, Texas, 1999.

LITERATURE CITED

[1] Klimentos, T., "Attenuation of P- and S-waves as a method of distinguishing gas and condensate from oil and water," Geophysics, Vol. 60, Society of Exploration Geophysicists, pp. 447-458, 1995.

[2] Rapoport, M. B., L. I. Rapoport, V. I. Ryjkov, V. E. Parnikel and V. A. Kately, "Method AVD (absorption and velocity dispersion) testing and using in oil deposit in Western Siberia," 56th Technical Meeting, European Association of Exploration Geophysicists, Extended Abstracts, Paper BO56, 1994.

[3] Dilay, A. and J. Eastwood, "Spectral analysis applied to seismic monitoring of thermal recovery," The Leading Edge, November 1995.

[4] Mitchell, J. T., "Energy absorption analysis: a case study," 66th Annual International Meeting, Society of Exploration Geophysicists, Expanded Abstracts, pp. 1785-1788, 1996.

[5] Batzle, M., D. H. Han and J. Cstagna, "Attenuation and velocity dispersion at seismic frequencies," 66th Annual International Meeting, Society of Exploration Geophysicists, Expanded Abstracts, pp. 1687-1690, 1996.

[6] Klimentos, T. and C. McCann, "Relationships among compressional wave attenuation, porosity, clay content and permeability in sandstones," Geophysics, Vol. 55, No. 8, pp. 998-1014, 1990.

[7] Martin, N. W., "Are P- and S-wave velocities and attenuations related to permeability? Ultrasonic seismic data for sandstone samples from the Writing-On-Stone Provincial Park in Alberta," MSc. thesis, The University of Calgary, 1996.

[8] Taner, M. T. and R. E. Sheriff, "Application of amplitude, frequency and other attributes to stratigraphic and hydrocarbon determination" in: "Seismic stratigraphy-applications to hydrocarbon exploration," The American Association of Petroleum Geologists, Tulsa, Oklahoma, pp. 301-327, 1977.

Nicolas Martin is a processing and research geophysicist with PDVSA E&P, S.A. and has worked within PDVSA since 1988. He earned a BS in physics at the Universidad Simon Bolivar;, Venezuela, in 1987, and MSc in geophysics at the University of Calgary, Alberta, Canada, in 1996. His areas of expertise are in seismic attributes, P-P/P-S seismic inversion and seismic modeling. He is a member of SEG, CSEG, EAGE and SOVG.

Armenio Azavache is an interpretive geophysicist with PDVSA E&P, S.A. He graduated from the Central University of Venezuela in 1974 with a geophysicist engineer degree, and began his career thereafter at the Ministerio de Minas e Hidrocarburos in Venezuela before coming to PDVSA in 1979. His expertise is in seismic interpretation. He is a member of SOVG and AAPG.

Maria Donati began work in 1988 at Seismoven SSC, S.A. before coming to PDVSA-Intevep, S.A. the following year, where she currently works as a research geophysicist. She earned a BS in physics at the Universidad Simon Bolivar in Venezuela and an MSc in geophysics at the University of Calgary, Alberta, Canada, where she is currently working on her PhD. Her expertise is in multicomponent seismic, A VO analysis, 3-D Pwave processing and multicomponent VSP. Professional memberships include SEG, CSEG, EAGE and SOVG.

COPYRIGHT 1999 Gulf Publishing Co.
COPYRIGHT 2000 Gale Group