Wellbore Path Planning Using ML Hazard Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The oil and gas industry faces challenges in minimizing time and cost associated with drilling wellbore paths to access hydrocarbon reservoirs, as existing methods lack effective predictive tools to avoid hazards and optimize wellbore paths based on geoscience and historical drilling data.
Innovation Solution
A method utilizing machine learning networks trained with geoscience and historical drilling data to predict drilling hazard probabilities along candidate wellbore paths, allowing for the determination of a safer and more efficient wellbore path to terminate at total depth coordinates within a hydrocarbon reservoir.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional drilling path planning methods are used, then drilling can proceed with existing procedures, but time and cost are increased due to lack of predictive hazard analysis
Solution Approach 1:
The machine learning model is trained in advance on historical drilling data and geoscience data to predict hazards before drilling begins. This preliminary action enables the system to provide predictive hazard probabilities and optimized path recommendations before the actual drilling operation, allowing operators to plan ahead and avoid hazards rather than reacting to them during drilling, thus reducing overall drilling time without requiring complex real-time analysis systems
Solution Approach 2:
The system uses historical drilling data from offset wells to create a predictive model that copies patterns of successful and unsuccessful drilling paths. By analyzing past drilling outcomes and creating a trained machine learning model that replicates this knowledge, the system can predict hazards for new well paths without requiring direct real-time measurement or complex physical analysis during drilling operations
2Reliability
If machine learning networks are trained with geoscience and historical drilling data to predict hazards, then drilling hazard probabilities can be predicted along candidate wellbore paths, but data processing complexity increases
Solution Approach 1:
The system merges multiple data sources including geoscience data, historical drilling data from offset wells, and candidate wellbore path information into a single machine learning model. By combining these diverse data types during the training phase, the model learns to integrate multiple factors that influence drilling hazards, improving prediction reliability while consolidating the complexity into a unified predictive system rather than requiring separate analysis systems for each data type
Solution Approach 2:
The machine learning model transforms raw geoscience and historical drilling data into predictive hazard probabilities by changing the parameters from raw measurements to normalized, standardized inputs that the model can process. This parameter transformation during training enables the system to handle diverse data types and scales, improving prediction accuracy while managing data processing complexity through systematic parameter standardization
3Productivity
If wellbore paths are optimized using predictive hazard analysis, then drilling safety and efficiency improve, but computational requirements increase
Solution Approach 1:
The computationally intensive machine learning model training is performed in advance using historical data and geoscience information. Once trained, the model can quickly evaluate candidate wellbore paths and predict hazards during the actual drilling planning phase. This preliminary computation shifts the heavy computational burden to before drilling begins, enabling fast, energy-efficient path evaluation during operations while still achieving optimized drilling paths that improve productivity
Solution Approach 2:
The trained machine learning model creates a compressed representation of vast amounts of historical drilling data and geoscience knowledge. This copied knowledge in the form of trained model parameters allows the system to make rapid predictions about drilling hazards without needing to re-process the original large datasets during each well path evaluation, significantly reducing computational energy requirements while maintaining high drilling efficiency
Data Source
AI summary
Methods and systems for wellbore path planning are disclosed. The method includes defining total depth coordinates of a candidate wellbore path within a hydrocarbon reservoir, obtaining geoscience data for a subterranean region enclosing the hydrocarbon reservoir, and obtaining historical drilling data from an offset well in the subterranean region. The method further includes training a machine learning network to predict drilling hazard probabilities along the candidate wellbore path using the geoscience data and the historical drilling data. The method still further includes determining a first wellbore path to terminate at the total depth coordinates using the candidate wellbore path and the trained machine learning network.