PATTERN RECOGNITION APPLIED TO THE INTERMITTENT GAS LIFT ONSHORE WELL CYCLING PROBLEM
Structural integrity; Oil & Gas; Artificial intelligence
The objective of this work is to develop a methodology to model the cycle times of the GLI (Intermittent Gas Lift) lifting method in onshore oil wells, using cycle adjustment data from the application of Artificial Neural Networks (ANNs) to optimize the gas injection used for artificial lifting, maintaining or increasing the oil recovery factor while meeting the mechanical strength criteria of the associated structures. Artificial intelligence has been used in a variety of applications in the engineering sector, especially in the oil and gas industry, and due to the complexity of the oil industry, the demand for new technologies has increased in recent years. The decision-making process for modeling cycle times proposed in this work can help reduce the frequency of operator visits to the well to check the flow produced within the cycle, as well as assist in the process of managing the structural integrity of production facilities in the oil and gas industry.