ML-Based Shaker Screen Selection for Drilling Solids
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Excessive low-gravity solids (LGS) in drilling fluids can adversely impact drilling operations by increasing the rate of penetration, equivalent circulation density, and surge/swap pressure, and making the filter cake thick and sticky, leading to equipment sticking issues.
Innovation Solution
A machine-learning model is used to predict particle size data from surface drilling, drilling fluid, and geological data, allowing for the determination of the appropriate shaker screen type and size to effectively remove solids, thereby optimizing drilling fluid properties and preventing equipment damage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a fixed shaker screen size is used throughout drilling operations, then equipment simplicity is maintained, but drilling efficiency decreases due to inability to adapt to varying formation characteristics and cuttings sizes
Solution Approach 1:
The system dynamically selects shaker screen sizes based on real-time drilling conditions, formation characteristics, and predicted cuttings size distributions. The machine learning model continuously processes drilling parameters and geological data to recommend optimal screen sizes, transforming the static shaker screen selection into a dynamic, adaptive process that responds to changing wellbore conditions
Solution Approach 2:
The patent replaces manual screen selection mechanics with an automated machine learning-based decision system. The ML model analyzes multiple parameters including drilling depth, formation type, rate of penetration, and hydraulic conditions to automatically determine optimal screen sizes, substituting human judgment and mechanical trial-and-error with computational intelligence
2Productivity
If larger shaker screen sizes are used, then equipment complexity is reduced, but solid removal efficiency decreases leading to excessive low-gravity solids in drilling fluid
Solution Approach 1:
The system applies different screen size specifications to different drilling conditions and formation types. Instead of using a uniform screen size across all operations, the ML model recommends specific screen sizes tailored to local conditions such as formation hardness, cuttings size distribution, and drilling fluid properties, optimizing solid removal efficiency for each specific drilling scenario
3Speed
If manual monitoring and adjustment of shaker screens is performed, then system simplicity is maintained, but response time to changing drilling conditions increases
Solution Approach 1:
The patent replaces manual monitoring and adjustment mechanisms with an automated machine learning system that continuously processes drilling data and automatically recommends screen size changes. The system substitutes human operators with computational algorithms that can analyze multiple parameters simultaneously and provide real-time recommendations, dramatically reducing response time to changing drilling conditions
Solution Approach 2:
The system implements a feedback loop where drilling parameters, formation data, and drilling fluid properties are continuously monitored and fed into the machine learning model. The model processes this feedback information and generates real-time recommendations for screen size adjustments, creating a closed-loop control system that adapts to changing conditions dynamically
4Reliability
If shaker screens are not optimized for specific cuttings sizes, then device complexity is minimized, but formation damage increases due to inadequate solid removal
Solution Approach 1:
The system performs preliminary analysis of formation characteristics and predicts cuttings size distributions before drilling operations begin or before transitioning to new formations. The machine learning model uses geological data and drilling parameters to anticipate solid removal requirements, allowing operators to pre-select appropriate screen sizes and prevent formation damage before it occurs
Data Source
AI summary
A method involves obtaining surface drilling data for a drilling operation at a wellbore, obtaining drilling fluid data regarding a drilling fluid in the wellbore during the drilling operation, obtaining drilling fluid hydraulic data regarding a drilling fluid device that causes the drilling fluid to circulate in the wellbore, obtaining geological data regarding one or more formations being traversed by the drilling operation, generating, predicted particle size data of cuttings in the drilling fluid using a machine-learning model, the surface drilling data, the drilling fluid data, the drilling fluid hydraulic data, and the geological data, and determining a shaker screen type based on the predicted particle size data. The method also involves transmitting a first command to a well control system, the first command being configured to change a first shaker screen to a second shaker screen in a shale shaker device based on the shaker screen type.