Virtual ESP Modeling for Early Wellbore Degradation Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging to accurately identify changes in wellbore operation conditions, such as those encountered in ESP operations, due to dynamic changes in system parameters, which can lead to incorrect conclusions about system conditions, and existing methods struggle to differentiate between changes in operating conditions and actual failures or degradations.
Innovation Solution
A statistical machine learning process is employed to monitor and analyze multiple wellbore variables over different time intervals, using paired random block design analysis and machine learning models to identify changes and determine root causes, allowing for comprehensive monitoring and detection of failures or degradations even when operating conditions change significantly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If visual inspection of individual parameters is used to monitor system condition, then the monitoring method is simple and easy to implement, but it leads to incorrect conclusions about system conditions due to dynamic operating changes
Solution Approach 1:
The patent combines multiple individual parameter monitoring into a unified system performance model that evaluates overall system condition. Instead of inspecting parameters separately, the invention integrates them into a comprehensive assessment framework that distinguishes operating condition changes from actual system degradation.
Solution Approach 2:
The invention transforms the monitoring approach from static parameter inspection to dynamic parameter analysis by establishing performance models that adapt to changing operating conditions. The system uses parameter relationships and trends rather than fixed thresholds to detect actual system changes.
2Ease of operation
If traditional monitoring methods are used, then the system is easier to operate, but it cannot differentiate between changes in operating conditions and actual failures or degradations
Solution Approach 1:
The patent segments the monitoring task into distinct components: operating condition analysis, system performance modeling, and degradation detection. By dividing the complex monitoring problem into manageable segments, the system can reliably differentiate between normal operating variations and actual system failures.
Solution Approach 2:
The invention introduces performance models as intermediary elements between raw parameter data and failure detection. These models act as mediators that process parameter relationships and trends, enabling reliable distinction between operating condition changes and actual system degradation.
3Measurement precision
If multiple wellbore variables are monitored using machine learning models, then the detection accuracy and root cause identification improve, but the system complexity increases
Solution Approach 1:
The patent develops universal machine learning models that can handle multiple wellbore variables and different time intervals simultaneously. These multi-functional models reduce system complexity by providing a single framework that performs detection, analysis, and root cause identification across various parameters rather than requiring separate analysis for each variable.
Data Source
AI summary
A method comprises selecting at least one wellbore variable of a wellbore operation and for each of two or more machine learning models, training each respective machine learning model with a respective set of training data values of the at least one wellbore variable detected in a respective different time interval. The method comprises for each of the two or more machine learning models, processing, for each of the at least one wellbore variable, a set of data samples of the respective at least one wellbore variable using each of the two or more trained machine learning models to output a respective model output response for each data sample of a set of data samples.