Automated computer vision system for real-time drilling cuttings monitoring. Han, R. Ph.D. Thesis, August, 2016. Accepted: 2019-02-01T23:51:29Z
Automated computer vision system for real-time drilling cuttings monitoring [link]Paper  doi  abstract   bibtex   
In rotary drilling operations, cuttings are continuously transported to the surface by drilling fluid. Real-time monitoring of cuttings and cavings is crucial for early detection and remediation of drilling problems such as stuck pipe, lost circulation, high torque and drag, reduction in rate of penetration, and other wellbore instability issues. These incidents are large contributors to drilling-related Non-Productive Time (NPT). At the current stage, a mud logger performs monitoring manually. This work proposes to use computer vision techniques to automate this procedure. To achieve this application, specific requirements should be established to design an automated machine vision system to maintain drilling safety and speed. Cuttings ramp has been identified as an ideal location to perform the measurement, where cuttings and caving are sliding down a slope at a steady speed. To accomplish this task, an intelligent image processing system must be able to track cuttings speed, measure volume, analyze size, and generate a surface model. Through a detailed review and testing of available 3D sensing techniques, a system consisting of a 2D high-resolution camera and 3D laser profile scanner was designed. By implementing image processing techniques, the cuttings speed on the ramp was estimated which was then synchronized to the 3D depth data from a laser scanner. Finally, the volume of moving cuttings was estimated and a 3D surface profile was reconstructed using point cloud data. Experimental results in the lab environment validated that such a system can be applied to quantify cuttings volume, size distribution, and reconstruct a 3D profile of cuttings and cavings. This measured result can be stored for further analysis. Overall, this work established a foundation for the design of a sophisticated real-time monitoring system for hole cleaning and wellbore risk reduction.
@phdthesis{han_automated_2016,
	type = {Thesis},
	title = {Automated computer vision system for real-time drilling cuttings monitoring},
	url = {https://repositories.lib.utexas.edu/handle/2152/72723},
	abstract = {In rotary drilling operations, cuttings are continuously transported to the surface by drilling fluid. Real-time monitoring of cuttings and cavings is crucial for early detection and remediation of drilling problems such as stuck pipe, lost circulation, high torque and drag, reduction in rate of penetration, and other wellbore instability issues. These incidents are large contributors to drilling-related Non-Productive Time (NPT). At the current stage, a mud logger performs monitoring manually. This work proposes to use computer vision techniques to automate this procedure. To achieve this application, specific requirements should be established to design an automated machine vision system to maintain drilling safety and speed.  Cuttings ramp has been identified as an ideal location to perform the measurement, where cuttings and caving are sliding down a slope at a steady speed. To accomplish this task, an intelligent image processing system must be able to track cuttings speed, measure volume, analyze size, and generate a surface model. Through a detailed review and testing of available 3D sensing techniques, a system consisting of a 2D high-resolution camera and 3D laser profile scanner was designed. By implementing image processing techniques, the cuttings speed on the ramp was estimated which was then synchronized to the 3D depth data from a laser scanner. Finally, the volume of moving cuttings was estimated and a 3D surface profile was reconstructed using point cloud data.  Experimental results in the lab environment validated that such a system can be applied to quantify cuttings volume, size distribution, and reconstruct a 3D profile of cuttings and cavings. This measured result can be stored for further analysis. Overall, this work established a foundation for the design of a sophisticated real-time monitoring system for hole cleaning and wellbore risk reduction.},
	language = {en},
	urldate = {2020-05-10},
	author = {Han, Runqi},
	month = aug,
	year = {2016},
	doi = {10.15781/T20G3HJ9R},
	note = {Accepted: 2019-02-01T23:51:29Z},
}

Downloads: 0