GOM Pre-Stack Inversion Course
Uploaded by
khj4qqg582GOM Pre-Stack Inversion Course
Uploaded by
khj4qqg582Common questions
Powered by AIThe non-uniqueness of pre-stack inversions implies that multiple geological models could explain the same seismic data. This poses challenges in accurately delineating subsurface features. In practice, this issue is addressed by incorporating geological constraints and cross-validation techniques, such as using well log data to establish a background trend between seismic variables like Zp, Zs, and density. These constraints help guide the inversion process towards more realistic models by identifying deviations from known trends, potentially signifying interesting geological features or fluid contents . Additionally, experts may iterate the inversion using different parameters and validate the results against independent geological knowledge.
The purpose of performing pre-stack seismic inversion using STRATA, particularly in the Gulf of Mexico example, is to convert seismic data into a more meaningful geological model. This process allows geoscientists to derive volumes of P-impedance, S-impedance, and density, offering the potential to distinguish between different subsurface materials and fluid contents . This information helps reduce the ambiguity in interpreting data by considering variations in seismic properties, which can indicate the presence and distribution of hydrocarbons in the subsurface .
Pre-stack inversion offers significant advantages over post-stack inversion primarily because it provides additional volumes that are critical for detailed geological and reservoir analysis. While post-stack inversion generally results in only P-impedance volumes, pre-stack inversion produces both S-impedance and density volumes alongside P-impedance. This comprehensive set allows for a more nuanced interpretation because it can be used to perform further analyses such as identifying fluid types and predicting lithological properties, thereby offering a richer understanding of the subsurface than is possible with post-stack inversion alone .
The background trend relating variables Zp, Zs, and density is crucial in pre-stack inversion analysis because it helps in reducing the non-uniqueness of the inversion results. In seismic inversion, multiple geological models can fit the input data, leading to ambiguities. By establishing a background trend based on well log data, often a linear relationship in the absence of hydrocarbons, geoscientists can detect deviations from these norms, which might indicate the presence of hydrocarbons or other geological features of interest . This approach enhances the reliability and interpretability of the inversion outputs.
The ZP and ZS cross plots help in understanding the relationship between different seismic variables in pre-stack inversion by graphically representing the correlation between acoustic impedance and shear impedance. These plots facilitate the detection of anomalies or trends that deviate from expected patterns, which might indicate geological features like hydrocarbon presence. The linear or near-linear relationships seen in the absence of hydrocarbons can be altered by reservoir conditions, thus providing insights into the subsurface geology. By analyzing these plots, geoscientists can better interpret the complex interrelations of seismic data, aiding in more accurate geological and reservoir characterization .
The conversion from offset to angle traces simplifies the workflow in pre-stack inversion by eliminating the need to perform this conversion as part of the initial data preprocessing steps in the workflow. By having this conversion pre-completed, as mentioned in the GOM V2 project example, users can skip certain steps, such as steps 2 and 3, thus allowing them to focus directly on subsequent processes, such as selecting appropriate volumes and horizons to use in the inversion . This streamlines the workflow and reduces the preparation time, enabling a more efficient inversion process.
Waveform expansion is crucial when viewing inversion results because it enhances the interpretability and clarity of detailed geological information. In STRATA, this is performed by expanding the display using specific icons which enable zooming and more detailed analysis of the Zp inversion result. This expanded view allows users to closely examine variations and features in the inversion outputs, facilitating better identification and understanding of structures like channels or faults at a finer scale . Enhanced visualization capabilities also assist in validating and refining interpretations by providing a clearer view of the impedance contrasts and other seismic attributes.
The inversion analysis approach starts by inverting a single Common Depth Point (CDP) at the well location to refine and verify the inversion parameters before expanding the process to the entire volume. This initial inversion serves as a localized test run, allowing fine-tuning of the model based on a well-controlled dataset where well logs and seismic data are optimally correlated. By addressing potential discrepancies at the well point, geoscientists can iteratively adjust the inversion parameters to improve model accuracy, which is then applied to the broader seismic volume with more predictable outcomes .
Extracting wavelets at different angle ranges in pre-stack seismic inversion is significant because it compensates for potential changes in seismic bandwidth across angles. Wavelets, which serve as the seismic signal's fingerprint, must accurately represent the true subsurface reflection characteristics. Different angles may display varying frequency bands, with further angles often showing reduced bandwidth compared to near angles. By using wavelets extracted from different angles, the inversion process better handles these variations, leading to more accurate and consistent modeling of seismic data .
In pre-stack seismic inversion, using different parameters like Delta ln(Zs) and prewhitening value plays a crucial role in optimizing the inversion results by controlling the balance between fitting the observed seismic data and maintaining a physically reasonable model. Different parameters can improve results in areas such as Vp/Vs ratios and minimize seismic residual errors. For instance, adjusting the prewhitening value can help reduce the influence of noise in the data, thus stabilizing the inversion process. Tailoring these parameters for specific datasets can enhance the accuracy and reliability of the resulting models, reflecting more accurately the underlying geology .