Geologic Volume Modeling with Local Conditioning Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current geological reservoir modeling techniques fail to accurately predict reservoir responses due to lack of conditionality and practicality in matching model and conditioning data, especially with vast model spaces and spatial heterogeneity.
Innovation Solution
A computer-method for modeling geological architecture by storing a geologic volume of interest model, obtaining local conditioning data, determining constraints based on this data, generating stochastic event models that conform to these constraints, and selecting and incorporating the most suitable models into the geologic volume of interest model, repeating this process for multiple flow events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional geological reservoir modeling techniques are used, then the model can be generated, but the model fails to accurately predict reservoir responses due to lack of conditionality and inability to match conditioning data
Solution Approach 1:
The patent segments the geological modeling process into discrete flow events occurring at different points in geologic time. Each flow event is modeled separately with its own stochastic generation, allowing systematic application of conditioning data at specific temporal and spatial locations. This segmentation enables the model to honor local conditioning data while maintaining overall geological realism.
Solution Approach 2:
The patent applies preliminary conditioning by first determining constraints from local conditioning data before generating stochastic realizations. The methodology pre-establishes boundary conditions and statistical parameters based on measured data (well cores, logs, seismic information) prior to model generation, ensuring that subsequent stochastic simulations are guided by actual observations rather than purely random processes.
2Adaptability or versatility
If the model attempts to match vast model space of potential heterogeneity results, then more comprehensive coverage is achieved, but it becomes impractical to expect coincidental model and conditioning data match
Solution Approach 1:
The patent applies local quality by making different regions of the model conform to different statistical distributions and conditioning constraints. Each flow event and spatial location can have unique heterogeneity characteristics derived from local conditioning data, rather than applying uniform statistics throughout the entire model volume. This allows the model to capture local geological variability while maintaining overall consistency.
Solution Approach 2:
The patent changes parameters by using stochastic simulation to generate multiple possible realizations of geological heterogeneity, then selecting or combining realizations that best match conditioning data. The methodology dynamically adjusts statistical parameters (variance, mean, correlation lengths) based on local conditioning information, allowing the model to adapt to observed data while exploring the vast model space of potential heterogeneity patterns.
3Reliability
If stochastic modeling is used to generate multiple event models, then the ability to conform to local conditioning data is improved, but the computational complexity increases
Solution Approach 1:
The patent applies partial action by generating a limited number of stochastic realizations for each flow event rather than exhaustively sampling the entire model space. The methodology generates enough realizations to ensure conformance to conditioning data (typically 1-3 realizations per event), then selects the best matching model, rather than generating and evaluating all possible heterogeneity patterns. This balances computational efficiency with reliability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A model of a geologic volume of interest that represents the geological architecture of the geologic volume of interest is generated. The model is generated as a series of geologic events at a string of points in geologic time such that each event is deposited or eroded sequentially. A given geologic event is determined based on the topological and/or geological properties of the geologic volume of interest at the time of the geologic event, environmental conditions present at the time of the geologic event that impact geologic formation, deposition, and/or erosion, and/or other considerations. The given geologic event is further determined to honor, at least somewhat, local conditioning data that has been obtained during direct measurements of the geological parameters (and/or trends therein) within the geologic volume of interest.