Quantitative interpretation of geophysical data requires rock physics as the bridge that links rock property heterogeneity observed at core and well scale to observations away from the well control. As wells become available, they can be incorporated into reservoir characterization studies for possible rock property calibration of zones of interests. If no other information is available, the modeling of probable reservoir quality zones can be done via pseudo-modeling exercises. This step might include empirical, theoretical or more sophisticated rock physics modeling techniques. It also assumes that an in-depth petrophysical model is available and proper well log data conditioning has been applied to the data.

The prediction and sensitivity analyses of rock properties are a fundamental phase in E&P and appraisal cases. It allows geoscientists to understand reservoir heterogeneity and possible facies scenarios due to changes in the elastic and electrical domains. By perturbing the rocks using a suitable rock physics model, explorationists can interpret key changes that might relate to a specific geophysical response.

Identifying a rock model accurate enough to represent rock type and its microstructure can be an exhausting task when performed in multiple reservoirs and wellbores. One common challenge is that teams will require constant input from rock physicists to first calibrate such a model and, most importantly, to constantly generate iterations that cover those possible scenarios that might explain the geophysical signature in question. The idea of a more interactive rock physics modeling approach is presented as a way to freeze the rock modeling phase and make it available to other geoscientists without generating unrealistic cases that are not supported by the rock physics diagnostics. It also allows interpreters to see these results in real time.

Multiwell studies
This concept can be applied to any size project, but it proves to be extremely convenient when dealing with multiwell studies or large areas including multiple well control points. Exploration efforts in areas such as the Gulf of Mexico (GoM), for example, have been ongoing for many years, and more areas have regained interest within and outside U.S. waters. Deepwater protraction areas such as Walker Ridge, Mississippi Canyon, Green Canyon, Alaminos Canyon and Keathley Canyon have continued to prove highly prospective over time. More than 540 wells have been included in rock physics studies; however, in the Alaminos Canyon area the interactive rock physics approach was applied to discovery wells drilled on the Trident, Silvertip and Great White prospects.

The modeling results for this example have been provided using a real-time modeling and visualization tool called rockAVO. Given the nature of these rocks, the base type of modeling is fluid substitution followed by calculation of the resulting synthetic seismic signatures. Fluid substitution modeling can provide a good understanding of amplitude vs. offset (AVO) response to fluid phase change for a given rock. In addition, matrix changes are performed so that the effect of changes in porosity and mineralogy of the reservoir can be observed instantaneously. Finally, fluid property modeling may be conducted to assess the AVO response to changes in oil gravity, gas gravity, gas-oil ratio, etc.

In summary, the overall objective of this approach is to encapsulate the underlying rock physics modeling methodology so users can interact with the data without the need to be a rock physics expert and without violating physical bounds defined during the rock physics diagnostics stage.

Workflow
The workflow is divided as follows:
• Step 1: All well data must be processed through a rigorous well log conditioning phase, which is termed Geophysical Well Log Analysis, and it includes the Rock Physics Diagnostics phase, which ensures that all logs are corrected (if needed) in a consistent manner. In this step, a rock model or combination of models are identified for reservoir quality rocks. This model will be used as a proxy for perturbational modeling purposes.
• Step 2: Based on the previous rock model, the reservoir is perturbed for variations in fluid, porosity and dominant mineral content as the main changing variables. Since the aim of the integrated result is an understanding of the theoretical geophysical responses to changes in reservoir properties, synthetic seismic and/or electromagnetic modeling also is incorporated into the workflow so that geoscientists can visualize the effect of changing these properties on post-stack and prestack seismic response and, ultimately, controlled-source electromagnetic data.
• Finally, all results are merged into the rockAVO visualization tool as a delivery mechanism, enabling users to effect and view changes in real time without changing the core rock physics of the methodology.

An example of the fluid sensitivity results for the main oil sand in the Great White well (AC857-1) is shown in the figure. The usage of interactive rock physics enables users to, for example, understand the theoretical variation of AVO signatures in the field using rock model constraints wrapped into the workflow panel. Quick quality control of AVO anomaly changes from Class III to Class II can be determined as a function of sand facies change. Lithoclasses in the area range from blocky oil sands to silty and shaly sands with residual oil saturations. Simultaneous changes to this particular example included API gravity, dissolved gas, water salinity and seismic geometry.

Rock modeling also can be used to build a template for efficient reservoir characterization so that numerous rock property scenarios can be modeled when interpreting geophysical data. This common technique allows us to understand reservoir property signatures with the principal objective of minimizing uncertainty and risk. For multiwell studies such as the GoM case, a rock physics template can be built upon the existing modeling conducted as part of the regional rock physics atlas for a specific location.

The panel presented in this case simplifies one of the many modeling combinations that can be presented in an interactive way to other geoscientists while keeping the rock model fundamentals untouched. It provides intuitive access to common rock physics practices that often are unavailable in multidisciplinary teams. The application of integrated workflows also ensures higher confidence in the modeling criteria when dealing with a larger number of well log datasets and a common source of rock physics modeling results.

In this dynamic deliverable in rockAVO for the AC857-1 well in Alaminos Canyon, elastic attributes (compressional or P-impedance, Poisson’s ratio and resistivity) are displayed after fluid substitution cases (80% gas in red, 80% oil in green and 100% brine in blue). From left to right, ray-traced synthetic gathers and stacks are shown for in situ, brine, oil and gas cases. The P-impedance and Poisson’s ratio plot (upper right) shows the upscaled response within the zone of interest and sand response (magenta square). The P-impedance vs. horizontal resistivity plot (lower left) shows the same sensitivity combining elastic and electric domains. The lower right plots show the reflectivity response at the depth indicated by the magenta line. The workflow panel to the right shows the parameter controls the user can interactively change while using rockAVO. (Source: Rock Solid Images)

References available. Contact Rhonda Duey for more information at rduey@hartenergy.com.