Loss Analysis and Diagnosis Based on Natural Language Processing and Big Data Analysis

The Block A in the western part of the Tarim Basin are mainly karst-vuggy and fractured reservoirs. Eighteen fault zones are developed in the block. The natural fractures located near the fault zones have complex distribution and low bearing capacity of the formation, which are prone to lost circulation. In order to accurately avoid the risk of lost circulation and optimize the technical measures to deal with the lost circulation, natural language processing technology was used to extract all the drilling and completion data and lost circulation information of Block A. Based on big data analysis, the uncertainty distribution of the equivalent density of the actual formation pressure and the actual fracture pressure in the leaky formation was summarized. The uncertainty range of fracture development and fracture width, as well as the lost circulation risk coefficient of the leaky formation were calculated, and the pre-drilling lost circulation risk diagnosis method was established. The case analysis showed that the proposed method could be used to diagnose the risk of lost circulation before drilling, which can provide a basis for avoiding the risk of lost circulation and developing the technical measures for lost circulation treatment during drilling and completion.