Study On Visual Laminar Flow Analysis Method Of Oil And Gas Wells

In the actual oilfield production and allocation process,production prediction is one of the key links.For different well conditions and well ages,production prediction is required as a prerequisite.Traditional production prediction needs to be predicted by numerical simulation or artificial intelligence.With the continuous development of horizontal well staged fracturing technology,unconventional oil and gas reservoirs such as low permeability reservoirs and shale reservoirs need to test the production of each section of horizontal wells in the actual production process.On the premise that the fracturing measures of each section are almost the same,the whole well production calculated by wellhead statistics cannot reach the theoretical value of construction.Therefore,the reservoir testing technology of liquid production profile in the process of staged fracturing has become a research hotspot.At present,the technology mainly includes production logging technology,distributed optical fiber technology and tracer monitoring technology.The three technical schemes have their own advantages and disadvantages,and cannot fully explain the distribution of multiphase flow.In view of this problem,the research work of this paper includes the following aspects:Firstly,based on the working principle of the visual detection system of oil and gas wells,a three-dimensional pipe string model is constructed in the Matlab simulation environment.The camera position is moved to simulate the working process of the Video Log downhole instrument scanning the wellbore,and the wellbore video image with laminar flow surface is collected.The laminar flow area ratio of the target image is calculated,and the correctness of the visual image laminar flow analysis method is verified by comparing the mathematical theory of the known laminar flow wellbore model.Secondly,the factors that may affect the acquisition of wellbore laminar flow video images are analyzed.In view of the key problem that the slicing method is selected according to the frame interval and depth interval in the manual processing process,which leads to the large error of the visual image laminar flow analysis results,by comparing the volume superposition results of 50 frame slices,100 frame slices and 200 frame slice intervals,a slicing method of first large cut and then small cut is proposed.Finally,aiming at the possible phenomena of fog and eccentric deformation in downhole video images,the dark channel defogging theory and image eccentricity correction method are studied and converted into actual processing algorithms by Open CV.Based on the real downhole video image of an shale well named Mou Dong,the laminar flow analysis of the d007 video is carried out,and the distribution and volume superposition of the target fluid in this section are counted,and the corresponding interpretation report is drawn.The experimental results show that the visual image laminar flow analysis method can grasp more comprehensive logging data by analyzing more effective information contained in the downhole video,which not only provides a verification method for other production prediction methods and evaluations,but also improves the oilfield development process,methods and strategies.It further provides scientific basis and reference value for improving oil recovery.