Abstract
Rate of Penetration (ROP) is one of the important factors influencing the drilling efficiency. Since cost recovery is an important bottom line in the drilling industry, optimizing ROP is essential to minimize the drilling operational cost and capital cost. Traditional the empirical models are not adaptive to new lithology changes and hence the predictive accuracy is low and subjective. With advancement in big data technologies, real- time data storage cost is lowered, and the availability of real-time data is enhanced. In this study, it is shown that optimization methods together with data models has immense potential in predicting ROP based on real time measurements on the rig. A machine learning based data model is developed by utilizing the offset vertical wells’ real time operational parameters while drilling. Data pre-processing methods and feature engineering methods modify the raw data into a processed data so that the model learns effectively from the inputs. A multi – layer back propagation neural network is developed, cross-validated and compared with field measurements and empirical models.
Highlights
The ever-increasing complexity of the wells escalate drilling cost and the focus of the drilling industry shifted towards reducing the NPT (Non-Productive time) and ILT (Invisible lost time)
In the prediction phase of machine learning process, the neural network model is applied on the drilling parameters of Well B which was drilled with the same drill string configuration to predict Rate of penetration (ROP) and compared with the empirical and regression model
It must be noted that the same feature engineering techniques applied on the Well A drilling parameters is applied on Well B drilling parameters before model is evaluated with the new data
Summary
The ever-increasing complexity of the wells escalate drilling cost and the focus of the drilling industry shifted towards reducing the NPT (Non-Productive time) and ILT (Invisible lost time). Rate of penetration (ROP) prediction with mathematical models and optimization of the drilling variables based on these models has been active area of research. ROP is the key business performance indicator in the drilling industry and a measure of additional contractual incentives Accurately optimizing it in real time with mathematical model has a direct business benefit for the industry. The abundance of real time drilling data together with increased big data infrastructure and computing resulted in the development of machine learning and non-linear statistical models for predicting ROP. A new model was developed to predict the ROP of adjacent wells by performing data analytics and developing an artificial neural network model from the offset well real time drilling data. Data Exploration The process of data exploration involves data selection, data quality analysis, exploratory data analysis and feature engineering
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Mathematical, Engineering and Management Sciences
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
