Enabling WITSML Stream Analytics Through Open Source Big Data Technologies

Abstract

This work presents a set of interconnected open source big data technologies through an example case to demonstrate the processes used to generate, process, store, and consume real-time wellsite information transfer standard markup language (WITSML) data from drilling and completions (D&C) operations. The new proposed approach to manage real-time data is based on the use of distributed storage and processing technologies to simultaneously analyze large volumes of information, previously considered only on an individual well basis. This new approach leverages the open source technology available in the market to target the gathering of the data, its secure database storage, and its future processing in a single workflow, to obtain the most value added from the oilfield-generated data during D&C operations. This work presents a set of big data tools and their application. For this case study, a dynamic, open, easily configurable, and fully scalable work environment was obtained. Open source big data technologies prove to be different when handling operations data, as compared to traditional technologies. Traditional technologies typically require several manual inputs and configurations to enter the data. Data models for these technologies are usually closed or proprietary, and the provided visualizations are limited either to single well analysis or to basic analytical dashboards. The technologies used for this case fully enable data input and storage, as well as allow the generation and real-time deployment of machine learning and data science algorithms and models. As a different approach, this paper introduces the use of big data architectures for the D&C business to obtain better results for gathering, storing, and analyzing real-time information than the business has traditionally achieved.