Bayesian Network Well Placement for Drilling Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional well placement techniques in the petroleum industry lack accuracy and efficiency in determining optimal locations for well drilling and lateral wellbore placement, often relying on outdated simulation methods that do not integrate multiple data types and require significant resources, leading to suboptimal hydrocarbon extraction.
Innovation Solution
A Bayesian Decision Network (BDN) expert system that analyzes geological and geophysical data to provide probabilistic assessments for well placement, completion type, and lateral wellbore direction, utilizing a combination of data availability, saturation, fracture, seismic, stress, thickness, porosity, tortuosity, washout, and net-to-gross values to generate precise probability scores for well productivity and completion recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional reservoir simulation and drilling technologies are used to determine well placement, then well production can be maximized, but the process requires significant resources and time
Solution Approach 1:
The patent creates a simplified probabilistic model that copies the essential assessment functions of complex reservoir simulation, allowing well placement evaluation to be performed through probability calculations rather than full-scale simulations. This reduces computational resources and time while maintaining decision-making quality
Solution Approach 2:
The system uses inexpensive, easily obtainable data sources (seismic data, well logs, production data) combined with rapid probabilistic algorithms to perform assessments that would traditionally require expensive, resource-intensive simulation models. The approach trades computational expense for speed and accessibility
2Measurement precision
If conventional techniques are used for well placement determination, then the process can be performed with existing methods, but accuracy and precision in identifying optimal locations are insufficient
Solution Approach 1:
The patent transforms the well placement assessment from deterministic simulation parameters to probabilistic parameters (probability values, confidence levels). This allows the system to express uncertainty and variability in predictions, improving accuracy by capturing the inherent unpredictability of subsurface conditions rather than providing false precision
Solution Approach 2:
The probabilistic assessment system serves multiple functions: it evaluates well placement options, assesses lateral wellbore directions, determines completion types, and identifies sweet spot zones. This multi-functionality improves measurement precision across different decision-making contexts without requiring separate specialized systems for each assessment type
3Reliability
If multiple data types are integrated for comprehensive assessment, then decision-making accuracy improves, but the system complexity and resource requirements increase
Solution Approach 1:
The patent merges multiple data types (seismic data, well logs, production data, geological models) into a unified probabilistic framework. Rather than analyzing each data type separately through complex specialized models, the system combines them into integrated probability assessments for well placement, lateral direction, and completion type, improving reliability while managing complexity through consolidation
Data Source
AI summary
A system for determining a well placement may include an analysis module. The analysis module, using a Bayesian Decision Network model, may perform a first assessment of data availability, saturation, fracture, and seismic data associated with a candidate area for drilling a first well, and may output a first probability that indicates a potential production level of the first well. The analysis module may perform a second assessment of offset, stress, thickness, and porosity data associated with a second, nearby well, and may output a second probability that indicates a potential production level of the first well if a lateral is placed in a layer and at a azimuth direction. The analysis module may perform a third assessment of tortuosity, washout, and porosity data, and a net-to-gross value associated with the first well and may output a third probability that indicates a completion type for the first well.