Font Size: a A A

Study On Casing Leak Detection Method Based On Signal Recognition Of The Collision Between Sand And Pipe Wall

Posted on:2018-12-22Degree:MasterType:Thesis
Country:ChinaCandidate:B LiuFull Text:PDF
GTID:2381330596468510Subject:Oil-Gas Well Engineering
Abstract/Summary:PDF Full Text Request
With the development of oil field,casing damage and sand control string become more and more serious.When the casing and sand control string damaged,excessive sand will affect the normal production of oil wells.It is important for casing maintenance work that the timely discovery and determine the depth of the leakage point.At present,the main methods used for casing damage detection in the world are mechanical caliper method,ultrasonic method,electromagnetic method and optical method.Each method has some defects,such as high cost,low detection efficiency or easy to be affected by the nature of the fluid in the casing.Therefore,a new method is put forward because the shortcomings.It can make the casing leak detection quickly and effectively,and provide technical support for safety production in oil field.This paper presents a leak detection method based on signal recognition of the collision between sand and pipe wall.This paper constructs a set of leak detection experimental device,and collect instrument,sensor and data processing method preferably,and designed leak detection experiment using the experimental device and through experiments the feasibility of the theory and method can be explored.During the experiment,the experiment uses the leak aperture and position as the experimental variables,analyzes signal characteristics,identifies sand impact signal,determines the impact time and at last calculates the leak location.Using wavelet packet analysis method to analyze experimental data,this paper finds the 16.2KHz~19.4KHz characteristic band,characteristics of sand impact signal obviously in the feature band.The results show that the experimental results coincided with the actual results,verify the feasibility of the prediction model of leakage depth.
Keywords/Search Tags:casing damage, signal detection, feature recognition, leak detection, wavelet packet analysis
PDF Full Text Request
Related items