Enhancing Plugging Material Selection in Deep Fractured Reservoirs with Quantum Neural Transformer Models

Effective fracture filling is essential for reducing fluid loss to protect deep naturally fractured reservoirs during drilling and completion. Controlling fluid loss in extensively fractured formations is a considerable challenge in drilling engineering during the exploration and extraction of deep subsurface resources such as oil, geothermal energy, and natural gas. If lost circulation is not managed, it may result in significant economic losses, prolonged non-production intervals, and safety issues. Moreover, in specific reservoir sections, significant fluid loss may hinder the exploration and extraction of abundant gas and oil resources. It is customary to incorporate LCM into drilling fluids to mitigate these challenges. This study primarily focuses on the selection of plugging materials using a QNN model. Data preprocessing approaches, including scaling and addressing missing values, were employed to improve the QNN model. PCA was subsequently utilized. The categorized impact of the QNN was employed to assist in selecting suitable filling materials for deep cracked reservoirs. The QNN model demonstrated superior performance compared to traditional models, with an accuracy rate of 98.26%. The findings indicate that the QNN model is viable and has the potential to enhance lost circulation management decisions while minimizing fluid loss during drilling and completion activities.