The Intelligent field is the oil industry's new trend that enables continuous monitoring and optimization of individual wells and overall reservoir performance. This is achieved by integrating fields' real time data in the reservoir management business processes. The results from this integration are anticipated to increase production rates, identify opportunities for higher hydrocarbon recoveries and reduce operating costs and future capital expenditures.Saudi Aramco has embarked on implementing the Intelligent Field (I-Field) initiative through new pilot projects in Qatif and Haradh increment III fields. The objectives of the pilot projects are to provide real-time diagnostic capabilities, highlight and address implementation challenges, and develop a comprehensive architecture for I-Field implementation in Saudi Aramco fields.This paper discusses the implementation approach of the intelligent field initiatives in Saudi Aramco. It will shed light on the challenges encountered and will present the process and methodology of developing the roadmap of the "surveillance layer," the first building block of Saudi Aramco's I-Field architecture.

The objective of this paper is to demonstrate the process of unleashing the potential of digital oil fields by combining the power of Big Data platform with the Internet of Things (IoT). This new method enables efficient machine learning training utilizing Big Data and real-time scoring against mathematical models for predicting future outcomes. Digital oil fields have a diverse set of IoT devices that measure important field metrics in real-time, such as downhole pressure, temperature and oil rate. A typical digital oil well is equipped with many equipment such as Multiphase Flow Meter, Electrical Submersible Pump, and Permanent Downhole Monitoring Systems. Those equipments have several sensors generating a huge volume of data every second. In order to enable data scientists to analyze this huge amount of data streaming from various data sources, a data engineering pipeline was built. This pipeline combines data from various real-time and historical data repositories along with a master relational database in order to provide a consistent and clean analytics database for data scientists. This method saves data scientists the trouble of manually preparing and cleaning data from different datasets. Furthermore, by utilizing the analytics database cluster for machine learning, data scientists were able to use bigger data sets for training their models which can improve the accuracies of the models. As part of the solution, a scoring engine was built which consumes real-time data feed from the digital oil fields and performs real-time predictions and scoring utilizing machine learning models. The new architecture significantly improved the productivity of data scientists by allowing them to focus on building models and not to have to worry about data plumbing and deployment of the model to the field. Moreover by utilizing bigger data sets, models accuracies was improved considerably. Finally by integrating the models with IoT real-time data stream, field engineers can see and act on the models’ predictions in a timely manner. This architecture and methodology combines different technology domains (IoT and Big Data) with unique solution to bring value to the Oil and Gas producing & production business function.

In this study, we aim to demonstrate how machine learning can empower computational models that can predict the flow rate of a given well. Given current real-time data and periodic well tests, this new method computes flow rates using data-driven model. The computational model is based on analyzing the relations and trends in historical data. Relational databases include huge amounts of data that have been accumulated throughout decades. In addition, there is a large number of incoming operational data points every second that gives a lot of insight about the current status, performance, and health of many wells. The project aims to utilize this data to predict the flow rate of a given well. A variety of well attributes serve as inputs to the computational models that find the current flow rate. Artificial Neural Networks (ANN) were used in order to build these computational models. In addition, a grid search algorithm was used to fine-tune the parameters for the ANN for every single well. Building a single unique model for every well yielded the most accurate results. Wells that are data-rich performed better than wells with insufficient data. To further enhance the accuracy of the models, models are retrained after every incoming patch of real-time data. This retraining calibrates the models to constantly represent the true well performance and predict better. In practice, Flow rate prediction is used by production engineers to analyze the performance of a given well and to accelerate the process of well test verification. One of the main challenges in building unique models for every well is fine-tuning the parameters for the artificial neural networks, which can be a computationally intensive task. Parameter fine-tuning hasn't been discussed in previous literature regarding flow rate prediction. Therefore, our unique approach addresses the individuality of every well and builds models accordingly. This high-level of customization addresses the problem of under-fitting in ANN well models.

No abstract