Using multidimensional analysis of historical drilling parameters combined with deep learning (DL) techniques, consistent ROP improvement in different drilling environments can be achieved. This discussion is focused on how offset well data can be properly analyzed and modeled to generate valuable outputs that improve not only drilling performance but also sustainability across the entire upstream. The workflow starts by history-matching comparable wells based on different variables such as well shape, hole size, wellbore design, downhole tools, etc. This first step is heavily based on the collaborative effect between subject matter experts working closely with data scientists to ensure only the right variables that are known to effect ROP are used and that a stable and scalable model can be aheived for different drilling environments. The process then moves to the data scientists where using multilayer perceptron models and random forest techniques allow determination of the ranking of features that affect ROP the most. The top tier features are then used to train a machine learning (ML) model to determine the average threshold of historic performance. Once the threshold is known, ML is again used to determine the optimal combination of drilling parameters that yield above-average ROP performance. This process is then repeated for each formation type and hole size. The performance range of the historical offset well data is then reviewed to determine the recommended threshold and fl score to output the highest modeling performance and the output parameter recommendations are then uploaded on a dashboard for real time guidance.

Recent developments in artificial intelligence (AI) have enabled upstream exploration and production companies to make better, faster and accurate decisions at any stage of well construction, while reducing operational expenditure and risk, increasing logistic efficiencies. The achieved optimization through digitization at the wellsite will significantly reduce the carbon emissions per well drilled when fully embraced by the industry. In addition, an industry pushed to drill in more challenging environments, they must embrace safer and more practical methods. An increase in prediction techniques, to generate synthetic formation evaluation wellbore logs, has unlocked the ability to implement a combination of predictive and prescriptive analytics with petrophysical and geochemical workflows in real time. The foundation of the real time automation is based on advanced machine learning (ML) techniques that are deployed via cloud connectivity. Three levels of logging precision are defined in the automated workflow based on the data inputs and machine learning models. The first level is the forecasting ahead of the bit that implements advanced machine learning using historical data, aiding proactive operational decisions. The second level has improved precision by incorporating real time drilling measurements and providing a credible contingency to for wellbore logging program. The last level incorporates petrophysical workflows and geochemical measurements to achieve the highest precision for logging prediction in the industry. Supervised and unsupervised machine learning models are presented to demonstrate the path for automation. Precision above 95% in the real time automated workflows was achieved with a combination of physics and advanced machine learning models. The automation of the workflow has assisted with optimization of logging programs utilizing technology with costly lost in hole charges and high rate of tool failures in offshore operations. The optimization has reduced the requirement for logistics associated with logging and eliminated the need for radioactive sources and lithium batteries. Highest precision in logging prediction has been achieved through an automated workflow for real time operations. In addition, the workflow can also be deployed with robotics technology to automate sample collection, leading to increased efficiencies.