Session: 11-10-01 Digitalization of Subsurface Wells Systems, Subsea Systems and Operations

Paper Number: 135747

135747 - Operations Optimization and Performance Analytics Powered by Machine Learning and Ai 

A typical Drilling Rig is equipped with hundreds of sensors which transmit data in real time (or near real time) which is then monitored by Drilling Engineers, Geologists and others in Real time Operation Centers to ensure a smooth operation and compliance to the plan. The drilling crews also report every day on the various aspects of the operation and a typical day/shift at a Rig starts with the analysis of the previous day’s reports. Extracting and interpreting information from morning reports and correlating it with the high frequency sensors data is complex and time consuming which makes it difficult to apply the learnings to control/improve/optimize the ongoing operation. This often results in inefficient operations with increased time, money and carbon footprint and can further cause safety incidents and accidents leading to the loss of lives. If the high frequency sensors data can be combined with the morning reports in time, it can highlight any potential problems/risks in advance and gives an opportunity to the Drilling crew to proactively prevent and/or mitigate the same. However, the large volume and frequency of data and the presence of multiple companies playing different roles on the Rig makes this task challenging, more so with the manual approaches often used currently. This case study will focus on a Smart performance analytics and optimization workflow enabled by an intelligent fusion of drilling sensors data with the daily operations reports and the BHA data, powered using a seamless integration of Drilling Domain knowledge with Machine Learning and AI techniques. It uses the learnings from historical data encapsulated into novel AI models which are then applied to real-time data so that the drilling engineers visualize, analyze, predict and optimize the various KPIs and operational parameters and further investigate potential areas of improvements to minimize ILT, NPT and carbon footprint of the operations.

Presenting Author: Sunil Garg DATAVEDIK

Presenting Author Biography: Sunil Garg (https://www.linkedin.com/in/sunil-garg-702409/) is the founder and CEO of dataVediK, a Houston based technology company building DataMoksha Hyper-Converged Data and Analytics Platform for Energy Industry using seamless integration of Energy Domain, Data Science and Software Engineering. Prior to this, he spent 20+ years establishing and growing Data Management, Big Data and Analytics business for Schlumberger. Sunil has deep understanding of Oil and Gas data, Data Science & ML and uses the combination to solve high uncertainty/high risk problems leading to seamlessly integrated end user centric solutions. He is a member of and an active participant in several industry organizations and forums like SPE, AAPG, PPDM, SPDM, OSDU and Rice Alliance. Sunil is a sought-after speaker at various industry conferences and conducts Big Data, Machine Learning and Blockchain trainings for the Industry, the Government and the Academia. He is also an active angel investor and a charter member of TiE, an organization whose mission is to foster entrepreneurship through mentoring, networking, funding and incubation.

Authors: