Multi-Attribute Seismic/Rock Physics Approach to Characterizing Fractured Reservoirs

Abstract

JUSTIFICATIONDuring Phase I, we performed our rock physics, geologic, and seismic feasibility analysis for fracture characterization. We addressed the problem of seismic fracture detection and characterization, in general, and specifically at our field site in West Texas. In this analysis, we were able to implement a number of fracture modeling procedures and to quantitatively assess the theoretical detectability of reservoir fractures. This allowed us to identify which of the several seismic attributes are most likely to be useful for fracture detection at the site. One of our results showed that Thomsen's parameters should be very useful for indicating gas-filled fractures at the site. We also outlined Monte Carlo approaches that can be applied for feasibility analysis at any site.Prior to our study, most seismic fracture detection methods have focused on qualitatively detecting seismic anisotropy. While anisotropy is critical, when used alone it has left us with interpretation ambiguities and nonuniqueness. During Phase 1 of this project, we found that by statistically integrating geologic and seismic information we can develop nontraditional methods for fracture detection. For example, we found that quantifying the correlations between fracture occurrence and depositional environment, and between fracture spacing and layer thickness, we allow us to decrease the uncertainty of fracture interpretation using seismic anisotropy. One of the reasons for this is that different depositional environments (and rock facies) give rise to differences in their seismic signatures (impedances, velocities, and Poisson's ratio.Phase II will allow us to validate and improve these methodologies by carrying out a smallscale field pilot study. Taking methods to the field, always allows us to make them more applicable. In the field, we will be able to more accurately assess the uncertainties introduced by measurement noise, source and receiver response, coupling, and near surface acoustic properties. We can then more realistically develop ways to reduce the uncertainty, as well as to identify which aspects of the geology and fracture distribution we can characterize most robustly. The small-scale field study will also help us to establish empirical fracture parameters that are needed as inputs to rock physics models. For example, ALL of the effective medium models for fracture-induced anisotropy require fracture stiffness as inputs (these are sometimes characterized as moduli or aspect ratios). There is no realistic way to assign these values DOE FINAL TECHNICAL REPORT ON PHASE 4 theoretically, so being able to calibrate them in a small pilot will give us the critical inputs we need for interpreting the full scale seismic survey in phase III. Finally, the data from the pilot will allow us to add an aspect of "data mining" to the study. In Phase I, we have exploited our best theoretical tools, but the real field data will give us the opportunity to discover other, unpredicted, signatures of fractures that might prove valuable.Our a...