Seismic Pore-Pressure Prediction Using Prestack Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Measuring pore pressure in low permeability rocks, such as unconventional reservoirs, is challenging due to poor vertical resolution of seismic velocities, which affects drilling and hydrocarbon production, and existing methods fail to accurately predict pore pressure changes with depth.
Innovation Solution
A method using seismic data inversion to generate pore-pressure transforms based on seismic impedance data, incorporating measured pore pressure data, upscaled sonic logs, and density logs, while adjusting for sampling bias and lithology variations, to produce accurate three-dimensional representations of pore pressure for improved prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If seismic velocity analysis methods such as normal move-out analysis and kinematic inversion are used, then pore pressure prediction can be obtained, but vertical resolution is poor and cannot resolve lithology changes over vertical depths
Solution Approach 1:
The patent transforms the input parameters from conventional seismic velocity data to seismic impedance data (P-impedance and S-impedance), which provides superior vertical resolution for detecting lithology changes. This parameter transformation enables the system to resolve pore pressure variations at different depths while maintaining sensitivity to reservoir property changes.
Solution Approach 2:
The patent introduces a pore-pressure transform as an intermediary computational framework that links seismic impedance data to pore pressure predictions. This transform incorporates measured pore pressure data, upscaled sonic logs, and density logs to bridge the gap between seismic observations and reservoir properties, achieving high vertical resolution without excessive complexity.
2Reliability
If a single seismic-velocity-to-pore-pressure transform is used, then the process is simple, but it fails to account for lithology changes and produces unphysical variations in predicted pore pressure
Solution Approach 1:
The patent applies different transform parameters specifically for shale lithology versus other formations. By identifying shale intervals and applying lithology-specific transforms, the system accounts for local variations in rock properties that affect pore pressure relationships, thereby improving prediction accuracy while managing model complexity through targeted differentiation.
Solution Approach 2:
The patent implements a dynamic pore-pressure transform that adjusts parameters based on depth and lithology type. The transform evolves from a simple single-parameter model to a depth-dependent, lithology-specific model that adapts to changing subsurface conditions, reducing unphysical variations while maintaining computational feasibility.
3Measurement precision
If pore pressure measurements are restricted to a subset of lithologies, then measurement complexity is reduced, but sampling bias occurs and prediction accuracy deteriorates
Solution Approach 1:
The patent incorporates measured pore pressure data from available lithologies as feedback to calibrate and validate the pore-pressure transform. By using these measurements to refine transform parameters and then applying the validated transform to predict pore pressure in unmeasured lithologies, the system compensates for sampling bias while maintaining practical data collection constraints.
Solution Approach 2:
The patent develops a universal pore-pressure transform that, once calibrated using data from a subset of lithologies, can be applied across multiple lithology types including shales and non-shale formations. This multi-functional transform reduces sampling bias by extrapolating from limited measurements while accounting for lithology-specific characteristics, balancing measurement representativeness with data collection feasibility.
Data Source
AI summary
A method of predicting pore pressure based on seismic data can include obtaining seismic inversion data based in part on seismic data collected from a formation. The method also includes calculating a pore-pressure transform, wherein the pore-pressure transform comprises parameters derived using measured pore pressure data, upscaled sonic logs, and density logs, wherein the pore-pressure transform comprises an objective function to reduce unphysical variations in predicted pore pressure corresponding to depth. Additionally, the method can include adjusting the pore-pressure transform for sampling bias caused by pore pressure measurements being restricted to a plurality of lithologies by accounting for a difference between upscaled seismic velocities and average sonic velocities within each of the lithologies. Furthermore, the method can include generating pore pressure prediction values based on the pore-pressure transform for the lithologies and the seismic inversion data, and modifying a seismic model based on the generated pore pressure prediction values.