Petrophysical Property Estimation Using Full Wavefield Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for characterizing subsurface rocks and hydrocarbon reservoirs face challenges in generating absolute petrophysical property estimates due to bandlimited seismic data, limited well data, and high uncertainty in petrophysical inversions, particularly in early stages of hydrocarbon exploration where well data is sparse.
Innovation Solution
A computer-implemented method using Full Wavefield Inversion (FWI) data and seismic angle stacks to iteratively determine low frequency components of rock and fluid parameters, employing a Bayesian approach with hidden Markov chains and forward-backward methods to estimate petrophysical properties and their uncertainties across multiple scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If bandlimited seismic data is used for petrophysical inversion, then the inversion process is computationally feasible and data acquisition is simpler, but absolute petrophysical property estimates cannot be generated and uncertainty is high
Solution Approach 1:
The patent applies asymmetry by treating low frequency and high frequency data differently. Full wavefield inversion results are used specifically to generate low frequency models, while seismic angle stacks provide high frequency constraints. This asymmetric division allows each data type to contribute its strengths without requiring complete spectral overlap.
Solution Approach 2:
The patent introduces low frequency models as an intermediary component that bridges the gap between bandlimited seismic data and absolute petrophysical property estimates. These low frequency models serve as a mediator that fills in the missing low frequency information, enabling the inversion to produce absolute rather than relative property estimates.
2Measurement precision
If detailed low frequency models are constructed to improve inversion accuracy, then absolute petrophysical property estimates can be generated, but the process complexity and computational requirements increase significantly
Solution Approach 1:
The patent applies preliminary action by generating low frequency models before the main petrophysical inversion process. These pre-computed low frequency models are then integrated into the inversion as constraints, simplifying the overall inversion process rather than requiring complex iterative low frequency modeling during the inversion itself.
Solution Approach 2:
The patent merges multiple data sources and modeling approaches into a unified inversion framework. Full wavefield inversion results, seismic angle stacks, and low frequency models are combined in an integrated inversion process that simultaneously utilizes all constraints to estimate petrophysical properties, reducing overall process complexity through consolidation.
3Reliability
If multiple spectrally variable datasets are integrated, then more comprehensive petrophysical property estimates can be obtained, but data processing and integration complexity increases
Solution Approach 1:
The patent applies local quality by allowing different datasets to contribute their specific spectral characteristics at appropriate frequency ranges. Seismic angle stacks constrain high frequency properties while full wavefield inversion results constrain low frequency properties, with each dataset optimized for its specific frequency domain rather than requiring uniform processing across all frequencies.
Solution Approach 2:
The patent utilizes parameter changes by transforming multiple spectrally variable datasets into a common parameter space suitable for joint inversion. Different datasets with varying spectral content are converted into consistent petrophysical property parameters, allowing integration while maintaining the unique spectral information from each source.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer-implemented method for determining rock and fluid parameters of a subsurface region from measured seismic reflection data, said method including: generating, with a computer, a geophysical data volume by combining a plurality of angle stacks obtained from the measured seismic reflection data and geophysical property data obtained from a full wavefield inversion of the measured seismic reflection data; for each point of the geophysical data volume, determining, with a computer, a petrophysical model that is a probability of a rock state based on initial values of the rock and fluid parameters and the geophysical data volume; iteratively determining, using a computer, updated values for the rock and fluid parameters, wherein the iteratively determining includes determining a petrophysical parameter estimate for the rock and fluid parameters from the petrophysical model as constrained by the geophysical data volume and the initial values of the rock and fluid parameters, minimizing a misfit between the geophysical data volume and synthetic data generated from a forward modeling of the initial values of the rock and fluid parameters using a cost function that includes the petrophysical parameter estimate of the rock and fluid parameters, and repeating the iteratively determining until a predetermined stopping criteria is satisfied and final values for the rock and fluid parameters are generated, and each subsequent iteration of the iteratively determining replaces the initial values for the rock and fluid parameters with the updated values for the rock and fluid parameters from a previous iteration; determining, with a computer, uncertainty in the final values for the rock and fluid parameters; and exploring for or producing hydrocarbons using the final values for the rock and fluid parameters and there uncertainty.