Download presentation
Presentation is loading. Please wait.
1
Digital Innovation in Oil & Gas
2
Reduced downtime/failures Artificial lift automation
Digital drives oilfield efficiency in different ways Digital investments can focus on surface equipment, wells, or the reservoir itself Surface Operations Equipment sensors Reduced downtime/failures Improved safety To Target Wells Downhole sensors Artificial lift automation Reduced lifting cost Daily optimization To Target The Reservoir Reduced lifting cost Optimized CAPEX Improved recovery Digital Oil Recovery™ To Target
3
Tubing Pres_slope >= -0.094
Case Study – Operations on the Surface A classification model used sensor data to predict spill events at a well and pad-level Objective Sensor data and tree-based modeling was used to develop a classification model to predict spills up to 7 days in advance Use well and pad-level sensor data to predict and avoid future events. Start with hundreds of spill events from dozens of wells and sensor data captured on the day of spill N= 1000 Tubing Pres_first.quantile >- 59 Oil Water Chemical N= 354 N= 646 Sensors used in models included casing, flowline, and tubing pressures Casing Pres_Peaks < 11 Tubing Pres_slope >= N= 472 N= 174 Tubing Pres_third.quantile < 63 Casing Pres_var >= 1.6 Operations Drilling Other Events were classified based on key behaviors observed in sensor data Development N= 410 Casing Pres_first.quantile >= 8.1 Technical Failure Undesired Behavior Other Results are used to develop classifier model to predict future spills based on sensor data Equipment & Tools N= 257 N= 97 N= 299 N= 111 N= 62 N= 118 N= 56
4
1-Day Ahead Well-Level Results 7-Day Ahead Well-Level Results
Case Study – Operations on the Surface Model accurately predicted future events The model predicted spills based on sensor data with an accuracy ranging from 77 – 88% 1-Day Ahead Well-Level Results 7-Day Ahead Well-Level Results Predicted Spill? Predicted Spill? No Yes No Yes No 64 11 No 72 6 Actual Spill? Actual Spill? Greg Mitchell Yes 3 22 Yes 5 17 True Positive Rate: When there is actually a spill, how often does the classifier predict yes? TP/actual + = 22/25 = 88% True Positive Rate: When there is actually a spill, how often does the classifier predict yes? TP/actual + = 17/22 = 77%
5
Reduced downtime/failures Artificial lift automation
Digital drives oilfield efficiency in different ways Digital investments can focus on surface equipment, wells, or the reservoir itself Surface Operations Equipment sensors Reduced downtime/failures Improved safety To Target Wells Downhole sensors Artificial lift automation Reduced lifting cost Daily optimization To Target The Reservoir Reduced lifting cost Optimized CAPEX Improved recovery Digital Oil Recovery™ To Target
6
Behavioral Forecast Model:
Digital Oil RecoveryTM Powered by FOROIL Machine learning, production data, and reservoir and well physics are used to construct a behavioral forecast model that accurately forecasts future production Gather measured production data Combine with equations of reservoir and well physics to generate a set of forecast models Scan the space using machine learning to find the best forecast model Best Forecast Model Model 1 Behavioral Forecast Model: +/- 5% Model 2 Model n
7
Digital Oil RecoveryTM Powered by FOROIL
The behavioral forecast model is combined with massive parallel computing to run 15 million potential field development plans overnight to optimize the reservoir Optimize across millions of development plans to find the optimal plan for the operator’s objective. Case Studies
8
Currently implementing
Future of Digital Technology in Oil & Gas Currently implementing Aspiring to long-term Exploring mid-term Well optimization 3D reservoir modelling Seismic data modeling and interpretation Digital twins Predictive maintenance Virtual reality surveillance Advanced process control Machine learning Prescriptive maintenance Integrated earth models Integrated remote operations centers Remote controlled robots GIS tracking for optimizing field workforce Automated sales marketing Autonomous drilling Automated well designs Well lifecycle analytics Digital procurement Integrated cross supplier ecosystem Digital-driven value pricing Automated 3D printing of parts Blockchain and smart contracts
9
Final Thoughts Still in the early-days of using new digital technologies to uncover opportunities for operations improvement There is a temptation to drive a digital agenda by data availability or computing power alone When opportunities are driven by business needs, targeted analytic techniques yield results that can be deployed Low CAPEX and high potential payoff means industry should increase pilots for targeted digital technologies in the field Workflows will change; organizational/operating model shifts will be needed to exploit the full value of digital
10
About Deloitte Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. Please see for a detailed description of DTTL and its member firms. Please see for a detailed description of the legal structure of Deloitte LLP and its subsidiaries. Certain services may not be available to attest clients under the rules and regulations of public accounting. Copyright © 2016 Deloitte Development LLC. All rights reserved. 36 USC Member of Deloitte Touche Tohmatsu Limited
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.