Impedance inversion and compressional (P) and shear (S) wave velocity ratio Vp/Vs estimations provide valuable information for improved reservoir characterization. At Rulison field, Colorado, the optimum reservoir quality sandstones can be characterized from the P- and S-wave seismic amplitude ratio volumes. The methodology produces a robust outcome, as seismic derived Vp/Vs is well correlated with log derived Vp/Vs. The velocity ratio is a determinant of reservoir rock quality in the tight gas sandstone environment. Investigated in laboratory, the core samples revealed that the presence of gas lowers the velocity ratio to the abnormal point, below 1.6. Our estimates of Vp/Vs from multicomponent seismic data confirm this result. Analysis of obtained Vp/Vs volumes in combination with the results from other independent studies performed on the field data evinces a good agreement between reservoir rock quality and velocity ratio values.

The production from unconventional gas reservoirs, such as Rulison field becomes very costly and challenging. Because of the reservoir heterogeneity, the exploitation approach chosen in the industry is to complete multiwells with dense spacing hoping that gas-bearing sandbodies can be found and produced. An often-encountered problem is establishing the sandbody connectivity from well to well even at 10 acre (660 ft) spacing. Another problem is determining what intervals are flowing and which are not as these wells are completed as commingled over the entire production interval. Multicomponent seismic brings the opportunity to analyze P-wave and S-wave type velocity. In this paper, we investigate velocity ratio Vp/Vs as an interpretation tool for improved reservoir characterization and lithology content discrimination in a tight gas sandstone environment. By estimating amplitude-derived values of Vp/Vs, we demonstrate their applicability for prospect identification and optimal well planning.

Methodology for Vp/Vs estimation

The flowchart used for velocity ratio Vp/Vs extraction from multicomponent seismic amplitudes is outlined in Figure 1. The inputs are 3-D zero offset stacked and migrated reflectivity data in the time domain of the registered wave types. The application of poststack seismic amplitudes assures high signal-to-noise ratio, and more reliable final estimations. Seismic inversion for impedance is the main procedure in the workflow. Since any inversion algorithm provides identical results within the seismic bandwidth, the choice of the inversion mechanism for Vp/Vs estimation is not substantial. We used model based inversion technique, which is available in Hampson & Russell. Low frequency subsurface models for P- and S-impedance were computed from a cross-dipole sonic log data available in a control well located within the study area (the information from this well is also used for seismic calibration, interpretation and investigation of the relationship between elastic properties of rocks and lithology). The inversion resulted in P- and S-impedance volumes generated from P-wave and either SS- or PS-wave data. These outputs are in the original time domain of the input seismic and contain a broadband frequency range. Before calculating the ratio of impedance volumes, they need to be transferred to the same vertical scale. We found it is better to perform this step using time rather than depthscale.

This content is only available via PDF.
You can access this article if you purchase or spend a download.