Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
A Case of a Class-4 AVO Anomaly In Gas Reservoirs
Xue, Lei (Liangbo, Sichuan Geophysical Company, PetroChinaChengdu) | Yalin, Li (Liangbo, Sichuan Geophysical Company, PetroChinaChengdu) | Minghua, Ouyang (Liangbo, Sichuan Geophysical Company, PetroChinaChengdu) | Yuan, Zhao (Liangbo, Sichuan Geophysical Company, PetroChinaChengdu)
Summary A comprehensive AVO analysis led to gas reservoir characterization of a tight clastic sedimentary rock in the upper Triassic, Sichuan Basin, China. A Class-4 anomaly has been identified by AVO numerical modeling using well logs. The modeled AVO anomaly was then verified by the amplitudes in the CIP gathers from prestack time migration. Based on cross-plots of 11 pairs of AVO attributes, three pairs --- intercept versus gradient, lambda-rho versus murho, and extended elastic impedance (near) versus extended elastic impedance (far) were selected as the most sensitive to the reservoir characteristics. The intercept versus gradient, and lambda-rho versus mu-rho cross-plots were used to characterize thick reservoirs with porosity greater than 8%. The extended elastic impedance (near) versus extended elastic impedance (far) was identified to be useful to distinguish qualitatively the difference in water saturation in the post-reservoir units. Eventually, we performed simultaneous inversion in order to obtain the attribute pairs (lamda-rho, mu-rho) and (extended elastic impedance-near and -far). As a result, we achieved a demonstrable AVO prediction which was verified by drilling. Introduction The target formation is an upper Triassic tight shaly sandstone with a depth of burial between 1,700-2,300 m and a thickness of 80-100 m. The average shale content is about 15% based on 46 wells in the 2D project (Figure 1). Within the shaly sand formation, there exist gas-bearing reservoirs formed by stratigraphic traps, and shale, shaly siltite, siltpelite and coal interbeddings with thicknesses mostly less than 2 m. Reservoir cores from two wells (Yang Jiajing et al., 2005) showed that the reservoir rocks within the compact sand have a character of low porosity and low permeability --- the average porosity is only 4% ,the maximum porosity is about 15%, while the producing layers have porosity more than 8%. Water saturation varies mostly from 30% to 80% within the reservoir rocks. Moreover, that locations and thicknesses of the reservoir units vary siginficantly in the lateral and vertical directions (Figure 2), and the maximum thickness of a single producing reservoir unit is about 21 m. The gas reservoir zone has already been confirmed by drilling before 2006. We had two objectives in mind for applying AVO analysis. First, we wanted to delineate the spatial extent and the overall thickness of the reservoir sequence with high porosity. Second, we wanted to map the variations in water saturation within the reservoir sequence, as there are some wells producing gas with water with significant variations in output. As of January, 2007, there are 45 wells with sonic logs and more than 30 wells with array sonic logs in the potential gas zone depicted in Figure 1. Within the reservoir zones, there exists a high correlation between the porosity in the cores and the sonic and density logs, and between the water saturation in the cores and that from log interpretation (Gan Xiu’e et al., 2005-2006). Since January, 2007, drilling new exploration wells continued, therefore, we were able to use the most recent well data to validate our predictions quickly.
- Geology > Rock Type > Sedimentary Rock > Clastic Rock > Mudrock > Shale (0.47)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock > Sandstone (0.34)
Summary We present a robust residual gather flattening technique based on trace correlations, which time-aligns coherent events across offsets or angles. We show how gathers to which we apply this method following migration and residual moveout corrections yield more coherent stacks, cleaner AVO results as well as more reliable velocity estimates. Introduction Well-aligned or “flat” events are key for high quality amplitude versus offset or amplitude versus angle fitting. In the context of marine 3D and 4D processing we have recently reported (MacKenzie et. al, 2004) on our efforts to improve gather quality with automated dense higher order moveout correction (including improved filtering methods). Here, we show how residual non-flatness of the data that cannot be further improved with the velocity picking tools, can nevertheless be corrected for prior to AVO. This gather pre-conditioning can be applied in a time-variant manner and is therefore able to correct for conflicting time-shifts. We also show that polarity reversal events are correctly flattened. Method The flattening technique is based on relative crosscorrelations between traces in a gather and a pilot trace. The choice of pilot trace is driven by signal to noise considerations and is generally chosen to be a partial stack of near-to-mid offset traces. Key elements of our algorithm are: ? A robust correlation technique which calculates time-shifts as a function of offset/angle and twoway time. ? The use of the absolute maximum of crosscorrelation in order to preserve polarity reversals. ? Automatic editing of outliers and filtering of timeshifts within and across gathers. Our method then consists of: ? Higher order RMO after migration. ? A two-pass residual gather flattening using a longwindow (200 ms) followed by a short-gated window (wavelength driven, but typically 40-60 ms). ? A user defined pilot trace generally chosen as a partial stack of near-to-mid traces. Synthetic Examples We start by showing three synthetic examples which highlight the robustness and limitations of the method. Figure 1 shows a single event (with added white noise) with a “wobbly” event. For correct alignment the choice of pilot trace is clearly key. For example, if the pilot is chosen as the full stack, the event will still be flattened but shifted to a slightly later time than that of the near traces. Figure 2 shows that our algorithm preserves polarity reversals. Here, estimating spatially consistent time-shifts across the gather is key as this downplays events at the locus of the polarity reversal where the signal to noise is poor. Figure 3 shows a synthetic with a combination of events with conflicting time-shifts and different AVO effects, including one polarity reversal. Here, we apply a cascaded approach using first a long time-window (200 ms) followed by a shorter cross-correlation analysis (60 ms) which is able to resolve the conflicting shifts necessary to align events. The cascaded technique is our preferred approach: we generally find it to give superior results to a one-pass short window method. In the following section we shall demonstrate this on real data.