Artificial intelligence-driven fuzzy logic approach for optimal well selection in gas lift design: A brown field case study
This study investigates the application of a Mamdani type-2 fuzzy inference system (FIS) for optimizing well selection in gas lift design within the HYSOL field of the Niger Delta, Nigeria. Gas lift is a widely used artificial lift method for enhancing oil production, particularly in wells with insu...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Results in Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2590123025000155 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study investigates the application of a Mamdani type-2 fuzzy inference system (FIS) for optimizing well selection in gas lift design within the HYSOL field of the Niger Delta, Nigeria. Gas lift is a widely used artificial lift method for enhancing oil production, particularly in wells with insufficient natural reservoir pressure. However, selecting suitable wells for gas lift is complex due to the inherent uncertainty and variability in petroleum production data. Traditional decision-making approaches often struggle with this uncertainty, leading to suboptimal choices that can reduce efficiency and increase operational costs. Given the need for a more flexible and accurate well selection process, the Mamdani type-2 fuzzy logic model was employed in this study to address the high levels of imprecision associated with key production parameters, including Gas-Oil Ratio (GOR), Barrel of Oil Produced per Day (BOPD), Basic Sediment and Water (BS&W), Barrel of Fluid Produced per Day (BFPD), Gas Produced (MCFPD), Total Gas-Liquid Ratio (TGLR), Tubing Pressure (TBG PR), and Manifold Pressure (MNFLD PR). By utilizing both upper and lower membership functions to define the Footprint of Uncertainty (FOU), the model effectively categorized wells into ''Excellent,'' ''Good,'' and ''Poor'' for gas lift suitability. Out of the initial fifty-nine wells, the system identified nineteen as excellent candidates, six as good, and twenty as poor, with thirteen wells excluded due to insufficient data. The Mamdani type-2 fuzzy inference system demonstrated robustness in managing complex decision-making processes under uncertainty, offering a flexible and scalable approach to optimize gas lift operations. This method provides a nuanced evaluation of well suitability, adapting to fluctuating reservoir conditions and operational challenges, and underscores the utility of fuzzy logic in petroleum engineering for handling large datasets and uncertain environments. |
|---|---|
| ISSN: | 2590-1230 |