Directional Drilling Parameter Estimation for Adaptive Well Path Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current directional drilling techniques face challenges in accurately steering along a pre-defined well path and achieving consistent wellbore quality due to sub-optimal steering control, which is influenced by unverified fudge factors and unknown bit-type and earth formation characteristics, leading to difficulties in computing optimal control actions for BHA-rock interaction.
Innovation Solution
A computer-implemented method and system for model-based parameter and state estimation using online and offline machine learning techniques to estimate BHA-rock interaction parameters, allowing for real-time adjustment of steering forces and weight on bit, and detection of formation changes, thereby improving control actions in directional drilling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional directional drilling techniques are used with unverified fudge factors and unknown bit-type characteristics, then the device complexity is reduced, but the steering control accuracy and well path precision deteriorate
Solution Approach 1:
The patent applies parameter changes by using machine learning models to dynamically estimate BHA-rock interaction parameters (such as side force coefficient, axial force coefficient, and moment coefficients) instead of using fixed, unverified fudge factors. These parameters are continuously updated based on real-time measurement data, allowing the control system to adapt to varying formation characteristics and bit conditions, thereby improving steering control accuracy without requiring complex manual calibration procedures
Solution Approach 2:
The system implements self-service through automated parameter estimation using machine learning algorithms that continuously learn from measurement data during drilling operations. The model automatically adjusts the BHA-rock interaction parameters based on observed drilling behavior, eliminating the need for manual intervention or verification of fudge factors by operators, thus improving accuracy while keeping the control system manageable
2Productivity
If real-time parameter estimation and adaptive control are implemented, then the productivity and wellbore quality are improved, but the use of energy and computational resources increase
Solution Approach 1:
The patent applies preliminary action by performing offline training of machine learning models using historical drilling data before actual drilling operations. This pre-training phase establishes baseline parameter estimates and model structures that can then be quickly applied during real-time operations, reducing the computational burden and energy consumption during critical drilling phases while still achieving high productivity through adaptive control
Solution Approach 2:
The system implements partial action by focusing computational resources on estimating only the critical BHA-rock interaction parameters that most significantly affect steering control, rather than attempting to model all drilling parameters. This selective parameter estimation approach maintains drilling efficiency by concentrating computational energy on the most impactful variables
3Extent of automation
If automated control actions are implemented based on estimated parameters, then the extent of automation is increased, but the reliability and robustness to unknown formations decrease
Solution Approach 1:
The patent implements feedback mechanisms where the machine learning model continuously receives measurement data from sensors during drilling operations and uses this information to update parameter estimates and adjust control actions. This closed-loop feedback system allows the automated control to adapt to unknown formation characteristics in real-time, improving both the extent of automation and the reliability of the control system when encountering unexpected geological conditions
Solution Approach 2:
The system applies dynamics by making the control parameters adaptive rather than static. The BHA-rock interaction parameters are continuously updated based on real-time measurement data, allowing the automated control system to dynamically respond to changing formation conditions. This dynamic adaptation enhances both automation level and reliability by enabling the system to handle unknown formations effectively
Data Source
AI summary
Examples of techniques for model-based parameter and state estimation for directional drilling in a wellbore operation are provided. In one example implementation according to aspects of the present disclosure, a computer-implemented method includes receiving, by a processing device, measurement data from the wellbore operation. The method further includes performing, by the processing device, an online estimation of at least one of a parameter to generate an estimated parameter and a state to generate an estimated state, the online estimation based at least in part on the measurement data. The method further includes generating, by the processing device, a control input to control an aspect in the wellbore operation based at least in part on the at least one of the estimated parameter and the estimated state. The method further includes executing a control action based on the control input to control the aspect of the wellbore operation.