Improved Predictions in Oil Operations Using Artificial Intelligent Techniques

Hassan, Amjed (King Fahd University of Petroleum & Minerals) | Al-Majed, Abdulaziz (King Fahd University of Petroleum & Minerals) | Mahmoud, Mohamed (King Fahd University of Petroleum & Minerals) | Elkatatny, Salaheldin (King Fahd University of Petroleum & Minerals) | Abdulraheem, Abdulazeez (King Fahd University of Petroleum & Minerals)



Oil is considered one of the main drivers that affects the world economy and a key factor in its continuous development. Several operations are used to ensure continues oil production, these operations include; exploration, drilling, production, and reservoir management. Numerous uncertainties and complexities are involved in those operations, which reduce the production performance and increase the operational cost.

Several attempts were reported to predict the performance of oil production systems using different approaches, including analytical and numerical methods. However, severe estimation errors and significant deviations were observed between the predicted results and actual field data. This could be due to the different assumptions used to simplify the problems. Therefore, searching for quick and rigorous models to evaluate the oil-production system and anticipate production problems is highly needed.

This paper presents a new application of artificial intelligent (AI) techniques to determine the efficiency of several operations including; drilling, production and reservoir performance. For each operation, the most common conditions were applied to develop and evaluate the model reliability. The developed models investigate the significance of different well and reservoir configurations on the system performance. Parameters such as, reservoir permeability, drainage size, wellbore completions, hydrocarbon production rate and choke performance were studied. The primary oil production and enhanced oil recovery (EOR) operations were considered as well as the stimulation processes. Actual data from several oil-fields were used to develop and validate the intelligent models.

The novelty of this paper is that the proposed models are reliable and outperform the current methods. This work introduces an effective approach for estimating the performance of oil production system and refine the current numerical or analytical models to improve the reservoir managements.