Reservoir Realization Subset Selection for Surrogate Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current field development plan evaluation methods for hydrocarbon reservoirs are inefficient due to the need for extensive and expensive reservoir simulations to account for geological and petro-physical uncertainties, leading to inaccurate risk assessments and high computational costs, especially when decision-makers' risk attitudes change.
Innovation Solution
A system and method that selects a representative subset of reservoir realizations to construct surrogates insensitive to risk variation, allowing for efficient construction and evaluation of field development plans without re-starting the process, even when the objective function changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive reservoir simulations are performed to account for geological uncertainties, then risk assessment accuracy is improved, but computational cost and time increase significantly
Solution Approach 1:
The patent segments the complete set of reservoir realizations into a representative subset that captures the essential geological uncertainty characteristics. This segmentation allows risk assessment to be performed on a manageable subset while maintaining accuracy, resolving the contradiction between comprehensive simulation and computational feasibility.
Solution Approach 2:
The patent creates surrogate models that replicate the behavior of full reservoir simulations. These surrogates are constructed from the representative subset of realizations and can be evaluated rapidly without re-running expensive simulations, enabling flexible risk assessment under different scenarios while avoiding repeated computational costs.
2Reliability
If a large set of reservoir realizations is evaluated to characterize uncertainty, then reliability of risk assessment is improved, but device complexity and computational resources required increase
Solution Approach 1:
The patent divides the large set of reservoir realizations into a representative subset that maintains the statistical characteristics of the full set. This segmentation reduces system complexity while preserving reliability by ensuring the subset adequately represents geological uncertainty without requiring evaluation of every possible realization.
3Productivity
If traditional RSM approaches aggregate risk measures across all realizations, then computational efficiency is improved, but measurement precision of specific realization responses is lost
Solution Approach 1:
The patent segments the analysis by constructing individual surrogate models for each realization in the representative subset rather than aggregating across all realizations. This allows efficient evaluation of each realization's specific response to decision variables while maintaining overall computational efficiency through the reduced subset size.
Solution Approach 2:
The patent applies partial action by evaluating only a representative subset of realizations rather than the complete set. This partial evaluation maintains measurement precision for the selected realizations while achieving sufficient productivity through the reduced scope, avoiding the need to process every possible realization.
Data Source
AI summary
A system, method and computer program product for assessing field development plans selected based on a stochastic response surface, preferably, for hydrocarbon reservoir production. Assessment begins by assessing uncertainty associated with multiple decision variable configurations. A subset of realizations is selected. An individual surrogate is constructed for each subset realization. A reduced representative realization subset is determined, where the reduced subset is representative of the behavior/performance of all realizations of decision variable configurations.