Changepoint Detection for Noise-Tolerant Drilling Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the hydrocarbon industry, real-time processing of data for accurate control of drilling operations is hindered by noisy data and the need for human intervention, which can lead to suboptimal drilling parameters and increased costs due to manual analysis challenges and fatigue.
Innovation Solution
Implementing a system that segments data streams to identify changepoints, allowing for real-time analysis and modeling of drilling parameters such as weight-on-bit and drill-bit rotational speed, using changepoint detection algorithms to automate the interpretation of sensor data and provide timely adjustments to optimize drilling operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human operators manually analyze sensor data to control hydrocarbon processes, then decisions can be made based on interpretation of raw data, but the ability to accurately process incoming data for real-time control is degraded due to fatigue, high workload, and difficulty in manually analyzing complex noisy data
Solution Approach 1:
The patent replaces the mechanical human cognitive system with an automated computational system. Sensors continuously monitor process parameters and feed data to a control system that automatically detects changepoints and adjusts drilling parameters, eliminating manual data analysis and decision-making. This substitution resolves the contradiction by providing accurate real-time processing without human fatigue or workload limitations.
Solution Approach 2:
The control system performs self-service by automatically detecting changepoints in sensor data and adjusting drilling parameters without human intervention. The system monitors its own operational data, identifies significant changes in the drilling environment, and autonomously optimizes parameters such as weight on bit and rotational speed, thereby eliminating the need for difficult manual analysis while maintaining high accuracy.
2Productivity
If forward looking models are used to control the automated industrial process, then real-time control is improved, but the complexity of determining models from prior processes, theory, and experiments increases
Solution Approach 1:
The patent segments the continuous sensor data stream into distinct phases using changepoint detection. Instead of requiring complex comprehensive models, the system divides the drilling process into discrete segments separated by identified changepoints, allowing for simpler localized control decisions within each segment while maintaining overall real-time control capability.
Solution Approach 2:
The system focuses on detecting only the most significant changes (changepoints) rather than attempting to model and respond to all variations in the drilling process. This partial action approach identifies critical transitions in drilling conditions and responds to those specifically, achieving effective real-time control without the complexity of comprehensive process modeling.
3Productivity
If data is continuously monitored to detect changepoints for real-time parameter adjustment, then drilling efficiency is optimized, but the noise in sensor data may have significant impact on the ability to accurately detect occurrences
Solution Approach 1:
The patent implements feedback mechanisms where the control system continuously monitors sensor data, detects changepoints, adjusts drilling parameters, and observes the resulting changes in sensor readings. This closed-loop feedback allows the system to distinguish true changepoints from noise by evaluating whether parameter adjustments produce expected responses, thereby improving detection accuracy while maintaining drilling efficiency.
Solution Approach 2:
The system dynamically adapts its detection thresholds and parameters based on current drilling conditions and data characteristics. By adjusting the sensitivity and criteria for changepoint detection in real-time, the system can maintain high accuracy in noisy environments, identifying true environmental changes while filtering out noise, thus optimizing both detection precision and drilling productivity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A changepoint detector for modeling data received from at least one sensor in a proces in the hydrocarbon industry. The data is segmented into a plurality of segments and fo each segment a model is assigned and the data corresponding to the segment fit to tha model. A plurality of segmentations are thus provided and these segmentations ar evaluated and assigned weights indicative of the fit of the models of the segmentation t the underlying data. The segmentation models are further used to calculate a result tha may be input to a process control program.