Identification Method of Tubing Leakage Based on PCA-SVM in Gas Well

Tubing leakage gas wells has always been a common problem in oil and gas production. Due to the complex and harsh working conditions of the downhole tubing, the tubing is usually corroded, cracked, perforated, etc. The sealing failure or performance degradation of the tubing will cause leakage, and the tubing leakage will directly threaten the production security of gas wells. In this paper, the experiment of the leakage point at different positions is carried out by using the visual tubing leakage experiment system. Multiple features are extracted and compared. Based on the features extracted from multiple samples, a tubing leakage identification model based on Principal Component Analysis - Support Vector Machine (PCA-SVM) is established. The accuracy rate is 98.33 %, which can effectively identify the tubing leakage, and is of great significance for ensuring the production safety of offshore gas wells.