Full-Field Analysis: A Machine Learning Approach


Abdullah Ahmed Alakeely







File Size:

82 MB

View File:

Access limit exceeded

Access Count:



The numerical reservoir simulation model is a stable among the tools used to address the problem of describing the fluid flow behavior of producing reservoir from wells. The downside of using these models is the inherent cost of development and deployment, especially for small fields, and resources restrained projects. The recent advances in machine learning methods in well testing domain, ignited an interest to bring these capabilities to reservoir simulation and management.

In this work, supervised and unsupervised machine learning techniques were applied to reservoir management tasks, that are usually dealt with using a numerical reservoir simulation model.

Starting with well level analysis, simulation of downhole pressure response in producing wells, as a function of production and /or injection rate, using only field data is investigated. The methods used include algorithms composed of feed forward, recurrent, and convolution layers. The same is used to simulate water cut response as a function of operational parameters. The results suggest that it is possible to generate accurate prediction using these techniques.

Extending the analysis to field level, we showed how using unsupervised learning techniques helped in guiding samples selection for training, then we applied generative modeling techniques using variational autoencoder to the problem of spatial control in the reservoir form data. We compared the performance of it to autoencoder, and other machine learning algorithms to predict multiphase production profiles form wells. Our investigation indicates that it can be done successfully in undrilled locations.

We further applied conditional variational autoencoder along with deep feature interpolation methods to generate novel simulations that are not available in the training data set, extending the range of available simulated profiles without additional training or the need to conduct numerical simulation runs.

Finally, we showed, using real field examples, how the methods developed can be used in predicting wells` production performance in the future, using past production data, and wellhead surface measurements, and control as an input. We applied this to single, two-, and three-phase flow examples successfully. We compared the liquid rate production performance to Gilbert correlation, a commonly used solution for this problem. We showed that using machine learning algorithms carries less uncertainty in cumulative production estimation, compared to Gilbert. The solutions carries implications for cost, and field data collections strategies.

The methodologies we employed assumes no reservoir model and are purely data driven.

Press the Back button in your browser.

Copyright 2021, Abdullah Ahmed Alakeely: Please note that the reports and theses are copyright to their original authors. Authors have given written permission for their work to be made available here. Readers who download reports from this site should honor the copyright of the original authors and may not copy or distribute the work further without the permission of the author, Abdullah Ahmed Alakeely.

Accessed by: ec2-3-236-18-161.compute-1.amazonaws.com (
Accessed: Monday 28th of November 2022 04:32:42 AM