Research of Defoaming Detection in Shale Foam Drainage Wells Based on Machine Vision

Abstract

Shale gas is generally faced with the problem of bottom-hole accumulation in the middle and late stage of exploitation. Foam drainage gas production technology is one of the means to solve the problem of bottom-hole accumulation and restore the normal production of gas Wells. When the foam mixture reaches the ground, it needs to be defoamed. At present, the defoaming effect is mainly evaluated by regular manual sampling from the separator sampling port. In order to reduce the work intensity, realize automatic detection and improve the detection accuracy, a set of anti-foam detection system based on machine vision was developed and field test was carried out. The results show that: The defoaming detection system meets the requirements of field conditions, and the system has strong stability and reliability; The defoaming detection model based on DeepLabV3+ has high accuracy, and the processed images are similar to the manual sampling results; The minimum difference between the foam height measurement system and the manual measurement is 0mm, the maximum difference is 8.60mm, and the data with the difference of 5mm and below account for 94.4%, that is, the accuracy rate reaches 94.4%. It is concluded that the development and test of the defoaming detection system has verified the feasibility of using machine vision technology for defoaming detection, and has practical significance for promoting the intelligent development of the defoaming detection process and improving the efficiency of shale gas extraction.