New artificial intelligence and big data analytics process to enhance non-metallic pipe deployments in digital oil fields using workflows for disparate data sets

New artificial intelligence and big data analytics process to enhance non-metallic pipe deployments in digital oil fields using workflows for disparate data sets

Machine learning

Carbon fibers

Gasoline

Big data

Offshore oil well production

Predictive analytics

Metals

Reinforced plastics

Automation

Decision trees

Steel fibers

Forestry

Data Analytics

The objective of this paper is to share the results and benefits from a new artificial intelligence and predictive data analytics process. This new process integrates both geoscience and engineering requirements to enhance non-metallic pipe deployments in the digital oil fields, using workflows for disparate data sets. It is based on interpreting large amounts of field data obtained from various engineering servers and geoscience data bases. A wide range of data-sets, such as various well stream variables and non-metallic technology operating conditions for hundreds of wells, and many different non-metallic pipe variations, are interpreted based on disparate workflows in the form of engineering decision trees. The outcome of the artificial intelligence process is to recommend the optimum non-metallic pipe technologies and generate an automated scope of work for optimal field results, and to predict the business-planning requirements for multiple fields at different operating conditions. Moreover, as part of this new process, well-stream data such as temperatures and pressures, H2S contents, gas oil ratios and produced gas conditions, as well as pipe size requirements and statistics are all deciphered for hundreds of wells in disparate workflows to determine the optimum non-metallic pipe technology required for any given set of field conditions, without the requirement of the human element. The preliminary workflows were stress tested successfully using representative wells with actual field data and the results are described in this paper. This new process was developed for several different types of non-metallic pipe technology, both rigid and flexible types, such as Reinforced Thermoplastic Pipe (RTP) both carbon-fiber and steel reinforced, Thermoplastic Composite Pipe (TCP), Reinforced Thermosetting Resin (RTR) and Internal Lining. The decision trees utilize the minimum and maximum allowable operating conditions for all the different technology types in different diameters, combinations and operating conditions against engineering parameters and statistics to forecast the optimum type and quantity of any non-metallic pipe technology for any given field, at any given time. The automation of this new process to electronically access existing big data servers to generate an automated scope of work and to supplement existing business-planning software significantly improves and optimizes the planning and implementation of non-metallic pipes. Furthermore, having an automated system significantly reduces the engineering time, while simultaneously reducing the error window, using artificial intelligence, allowing engineers to spend more time addressing other pressing challenges. The new process also results in cost savings by ensuring fit for purpose applications and ensuring optimal designs are applied. Finally, the developed workflows also serve as a guideline for the utilization of surface nonmetallic pipe technology for any field conditions. This serves as a valuable reference to add to both the industry and the existing body of literature. 2020, Society of Petroleum Engineers

Alrabeh, Majed

Abuzaid, Abdulrahman

Abu Dhabi International Petroleum Exhibition and Conference 2020, ADIP 2020, November 9, 2020 - November 12, 2020

2020

conferencePaper