Well Placement Optimization Using Binary Matrix Convolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optimal well placement in complex and heterogeneous petroleum reservoirs is challenging due to uncertainties in reservoir flow properties, complex and multi-modal objective functions, and large optimization problem spaces, especially when considering constraints like well separation distance and rock quality.
Innovation Solution
The method involves converting constraint information into binary matrices, convolving them to determine feasible regions, and using a depth-first search algorithm to identify optimal well locations within these regions, iteratively updating the constraint matrix and feasible regions to optimize well placement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual determination of well locations is used, then simplicity and ease of operation are maintained, but optimization precision and recovery efficiency deteriorate
Solution Approach 1:
The patent segments the continuous reservoir space into discrete grid blocks, transforming the continuous well placement problem into a discrete optimization problem. This segmentation enables the use of binary matrices to represent feasible and infeasible regions, making the problem computationally tractable while maintaining optimization precision.
Solution Approach 2:
The patent introduces binary matrices as an intermediary representation between the physical reservoir and the optimization algorithm. These matrices encode constraint information and feasible regions, serving as a bridge that enables automated optimization while handling complex constraints systematically.
2Manufacturing precision
If automated optimization tools are used, then optimization precision improves, but computational cost and time consumption increase
Solution Approach 1:
The patent performs preliminary action by pre-processing constraint information into binary matrices before the main optimization process. Feasible regions are identified and encoded in advance, which significantly reduces the search space and computational burden during the actual optimization iterations.
Solution Approach 2:
The patent applies local quality by focusing computational efforts only on feasible regions identified through binary matrix representation. Instead of searching the entire reservoir space, the optimization algorithm concentrates on areas that satisfy constraints, improving computational efficiency without sacrificing optimization precision.
3Reliability
If constraints are strictly enforced, then solution feasibility improves, but the search space and optimization complexity increase
Solution Approach 1:
The patent transforms constraint parameters into binary matrix representations, where each cell indicates whether a location satisfies the constraint. This parameter transformation simplifies constraint handling by converting continuous or complex constraint conditions into discrete binary values that are easy to process and combine.
Solution Approach 2:
The patent merges multiple constraint representations into a unified binary matrix structure. By combining individual constraint matrices through logical operations, the system handles multiple constraints simultaneously in a systematic way, improving reliability without proportionally increasing complexity.
4Area of stationary object
If the optimization problem space is expanded to cover the entire reservoir, then comprehensive coverage improves, but computational burden increases drastically
Solution Approach 1:
The patent extracts and isolates feasible regions from the entire reservoir using binary matrix representation. By taking out only the relevant feasible areas and representing them compactly in binary matrices, the system reduces the effective search space while maintaining comprehensive coverage of all viable well locations.
Solution Approach 2:
The patent adds a binary dimension to the spatial representation of the reservoir. By introducing binary matrices that encode feasibility information, the system transforms the problem from a purely spatial search to a combined spatial-binary optimization, enabling efficient filtering and reduction of the search space.
Data Source
AI summary
Information associated with a plurality of constraints for well placement is received. For each of the plurality of constraints, information associated with the constraint is converted into a binary matrix. An overall constraint matrix is determined by convolving the binary matrices associated with the plurality of constraints. Feasible regions are determined based on the overall constraint matrix. Optimal well locations are determined within the feasible regions. The optimal well locations and a number of the optimal wells are output.