An Efficient AI Model to Identify Loss Circulation Zones in Real Time

Hossain, M. Enamul (Nazarbayev University) | Gharbi, S. H. (King Fahd University of Petroleum & Minerals) | Abduljabbar, A. M. (King Fahd University of Petroleum & Minerals) | Al-Rubaii, M. (King Fahd University of Petroleum & Minerals)



The drilling industry is going to face challenges due to lack of manpower, and new operational hazard in near future. In addition, drilling wells are also moving toward new and challenging operations such as deep water, shale oil, and harsh environment. Another difficulty to make the situation more difficult is that huge number of drilling experts are retiring from the industry soon, and they are going to be replaced by new, young, unexperienced engineers. The industry need to develop unconventional solution to overcome this situation. Some operation centres such as real-time operation centers and the geosteering operation centers can help. However, due to the human capabilities, these centers can handle a small fraction of the total drilling operations. One of the solutions is to utilize the computational power to develop artificial intelligent (AI) models that assist the drilling engineers and operational crews.

This paper discussed the development of an AI model which identify the loss circulation incidents. The model identifies these incidents in its early symptoms, before it matures to well stability problem or well control situation. In addition, it compares the current loss circulation identification methodology, and highlight how this model was successfully able to identify same event in advance, providing the drilling engineers, operation crews and drilling fluid specialist with bigger window to mitigate the situation, and resolve it in its early stages. Moreover, the paper discussed how integrating such model with more advanced hydraulic analyses concepts can lead to more sophisticated well control detector environment, or event fast formation top identifier.

The paper pointed that the AI are widely used in different disciplines while in drilling industry it is still crawling. It is very important for the drilling industry to invest in developing more advanced AI which can assist in predicting troubles and optimize drilling operations. If these models are developed, they can open new avenues such as automation in drilling in the drilling industry where one day an AI can handle the drilling operations in the seabed of the deep ocean.