Geobody Calculation Algorithm for Reservoir Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for calculating connected geobodies in 3D geological models of hydrocarbon reservoirs are computationally costly due to repeated visits to each cell and redundant labeling, making them inefficient for large models and impractical for analyzing multiple realizations within a realistic time frame using conventional computer hardware.
Innovation Solution
A method that assigns a unique geobody identifier to each cell by determining threshold criteria, identifying neighboring cells, and using a temporary store to ensure each cell is processed only once, allowing for efficient geobody calculation without excessive CPU time, applicable to various grid structures including Cartesian, corner point, and unstructured grids.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional geobody calculation algorithms (cluster multiple labeling, X-stack/Y-stack/Z-stack scanning, directional search) are used to identify connected geobodies in 3D reservoir models, then geobody connectivity can be determined, but CPU time becomes excessively costly for large models with one million or more cells
Solution Approach 1:
The patent applies preliminary action by first assigning a 'met flag' to each cell based on threshold criteria before the geobody identification process. This pre-processing step organizes the data structure in advance, allowing the subsequent breadth-first search to efficiently traverse only relevant cells without redundant operations, thereby reducing CPU time for large models
Solution Approach 2:
The patent extracts and removes redundant operations from traditional algorithms by using a temporary store to track visited cells. Each cell is processed exactly once during the breadth-first search, eliminating the repeated visits and redundant labeling that plague cluster multiple labeling and stack-based approaches, thus significantly reducing computational time
2Reliability
If algorithms that repeatedly scan and label cells (such as geo_obj and Petrel utility function) are used to ensure accurate geobody identification, then connectivity can be determined, but the processing time increases prohibitively for models with 50 million cells
Solution Approach 1:
The patent implements self-service through the temporary store mechanism that automatically tracks which cells have been visited and assigned geobody identifiers. Each cell 'services itself' by checking the temporary store before processing, preventing redundant operations without requiring external coordination or multiple scanning passes, thereby maintaining accuracy while dramatically improving processing speed for large models
Solution Approach 2:
The patent ensures continuity of useful action by implementing a breadth-first search that continuously processes cells in the temporary store without interruption or repeated scanning. The algorithm maintains a continuous flow of geobody identification through the model, assigning identifiers in a single pass rather than requiring multiple sequential scans, thus maximizing productivity while ensuring reliable identification
3Measurement precision
If directional search or iterative search methods are used to calculate geobodies from well perforation locations, then connected regions can be identified, but isolated geobodies not connected to current wells cannot be found and CPU time remains excessive
Solution Approach 1:
The patent applies universality by implementing a breadth-first search algorithm that can identify all geobodies in the model regardless of their connection to wells. The temporary store-based approach universally processes all cells with met flags, making the algorithm adaptable to find both well-connected and isolated geobodies, unlike directional or iterative search methods that are limited to well-perforation starting points
Data Source
AI summary
A method for calculating connected productive regions (“geobodies”) in a hydrocarbon reservoir uses an algorithm which visits a cell in a geological model of the reservoir, assigning a geobody identifier if the cell possesses certain threshold characteristics and adding the connected neighbors of each visited cell to a data queue or stack. The data queue is emptied cell by cell, each cell popped out of the queue being assigned the same geobody identifier value and having its own connected neighbors added to the queue. Cells are popped out until the queue is empty and then another cell in the model which has not yet been visited is selected. The process continues until all cells have been visited. Cells may also be pre-identified as being part of a well; in this case the algorithm will also establish well connectivity.