Facies Realizations Using Geobody Reassignment for Target Proportions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing facies probability cubes in reservoir modeling are fuzzy, leading to highly variable geostatistical realizations that poorly represent the subsurface volume of interest due to overlapping seismic property distributions of different facies, which complicates decision-making in reservoir management.
Innovation Solution
Integrate seismic data to generate facies realizations that preserve patterns and variability by using a geobody index to assign facies based on facies probability vectors, adjusting geobody assignments to match target facies proportions through random sampling and reassignment processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If facies probability cubes are generated using conventional methods, then the modeling process can be completed, but the probabilities are fuzzy (not near 0 or 1) leading to highly variable geostatistical realizations
Solution Approach 1:
The patent changes the parameter approach by transitioning from probability values (0-1 range) to frequency counts (0-N range). By counting the number of times each facies is selected during multiple simulation realizations, the output becomes discrete frequencies that can be directly interpreted as deterministic facies assignments, eliminating the fuzziness of probability values near 0.5.
Solution Approach 2:
The patent uses multiple copies of simulation realizations to build up frequency counts. By running the geostatistical simulation multiple times and counting facies occurrences across these copies, the method transforms stochastic probability outputs into deterministic frequency-based facies assignments, achieving clearer facies representation.
2Productivity
If geostatistical realizations are generated from fuzzy probability cubes, then realizations can be produced, but the representations are highly variable and poor representations of the actual subsurface volume
Solution Approach 1:
The patent implements feedback by counting the frequency of facies selections across multiple realizations and using these frequencies to determine final facies assignments. This feedback mechanism ensures that the most frequently selected facies (highest frequency count) are assigned to each location, producing consistent and reliable subsurface representations rather than highly variable ones.
Solution Approach 2:
By changing from probability parameters to frequency count parameters, the method transforms the output distribution from continuous probabilities to discrete frequency integers. This parameter transformation ensures that facies assignments are based on dominant patterns across multiple realizations, improving representation accuracy while maintaining productivity.
3Adaptability or versatility
If seismic property distributions of different facies overlap, then seismic data can be used for modeling, but the overlapping distributions cause fuzzy probabilities that complicate decision-making
Solution Approach 1:
The patent uses multiple copies of simulation realizations to overcome the ambiguity caused by overlapping seismic property distributions. By running simulations multiple times and counting facies frequencies across these copies, the method transforms ambiguous probability values into clear frequency-based facies assignments, making decision-making easier while still utilizing seismic data for modeling.
Solution Approach 2:
The patent changes the parameter representation from probability (0-1) to frequency count (0-N), transforming the output of models that deal with overlapping seismic distributions. This parameter transformation converts ambiguous intermediate probability values into discrete, interpretable frequency counts that directly indicate the most likely facies, simplifying decision-making while maintaining adaptability to seismic data.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods are disclosed for generating a set of facies realizations. A computer-implemented method may use a computer system that includes a physical computer processor and data storage. The computer-implemented method may include: obtaining a geobody index, obtaining facies probability vectors for the multiple geobodies, assigning facies to the multiple geobodies based on the facies probability vectors, obtaining a target facies proportion for the subsurface volume of interest, reassigning a first geobody having a first facies based on a first facies probability vector of the first geobody, and reassigning remaining ones of the multiple geobodies with different facies.