The Methods Research Of Denoising Of Microseismic Data

In recent years, oil supply increased year by year, the reservoirs low-permeability reservoir is gradually increasing in the proportion in the entire field of oil production. Hydraulic fracturing can transform stratum structure between the cbsed reservoirs and the wellbore had fractured out a new crack as a new fluid channel This method can greatly improve the oil production of low permeability reservoir reservoirs, acts as an indispensable role in the current exploration and development of oil and gas field. In the hydraulic fracturing process, we need to use the micro-seismic monitoring methods observing the spatial image of cracks of formation in the fracturing process, and to make a realistic evaluation of the fracturing effect of hydraulic fracturing. Important factors that affect the evaluation results, The level of the signal-to-noise ratio of micro-seismic data obtained after de-noising. Because the short duration of micro-seismic events, high frequency of sound waves, and the features of releasing energy smaller, collected micro-seismic data had mingled many interference noise signal in the actual production process, and some even interfering noise signals totally covered the active signal. Therefore, the collected micro-seismic signals to do de-noising processing before explaining, it is an urgent problem. Based on the type of micro to seismic signal interference noise signal is analyzed, combined with technical experience in terms of micro-seismic signal de-noising at home and abroad in recent years, analyzed and summarized in a single de-noising method, and the improved noise method is applied to the instance of the micro-seismic signal de-noising.This article from the actual production of the current field quo leads the importance of micro-seismic monitoring in the mid to late of oil exploration, a brief introduction abroad micro-seismic monitoring methods from proposed the theory to apply the actual production process, the methods of using detector collection the date of micro-seismic monitoring (typically well monitoring), micro-seismic monitoring system by which the instrument components. From the types and characteristics of the interference signal of micro-seismic data to start, mainly studied the predictive de-convolution filtering, correlation filtering, variable mode decomposition, Curvelet transform micro-seismic signalde-noising methods and so onFinally, the use of MATLAB compiled the programs of predictive de-convolution filter, band-pass filtering, correlation filtering, variable mode decomposition, Curvelet transform and so on Several of the previously mentioned reconstruction or filter de-noising for micro-seismic data of collected in the field is analyzed and processed in MATLAB language platform. Results of obtained show that:(1) In the actual production process, automatic picking of effective signal and real-time accurate imaging of crack position is closely related to the signal-to-noise ratio of micro-seismic data. (2) The position of the focal of micro-seismic signals to determine what the point is an important question of de-noising process. In the micro-seismic signal source location is unknown, many of the perfect application of seismic data processing methods are extremely limited, can not play its real role. (3) the micro-seismic data contains significant wave and interference waves, broken down into direct wave, refracted waves, wave transmission, multiple waves, converted waves, surface waves, pure guided wave and tube wave; longitudinal and shear waves for linear polarization waves, surface waves, guided waves for nonlinear polarization waves. (4) Principal component analysis of micro-seismic signals suppress random noise interference have some effect, the effect of noise suppression better than the band-pass filtering. (5) The frequency of the filter is based on this feature of the effective signal-to-noise separability in the frequency domain, setting a signal to suppress a particular frequency band which played the effect of nose suppression. Therefore, the use condition of the frequency domain filter is only when there is a big difference between valid signal and noise in the frequency domain in order to play better suppression noise. (6) Predictive de-convolution filtering can compress wavelet and effectively remove the multiple waves in micro-seismic data. Predictive de-convolution filter does not need auxiliary information such as speed and running speed. (7)Correlative filter is an effective method on the background of random noise emergent the axis of micro-seismic signals. Micro-seismic data based on the measured cross-correlation parameter selection is reasonable, can effectively remove the random noise interference, improve phase axis resolution. Micro-seismic data based on the measured cross-correlation parameter selection is reasonable, can effectively remove the random noise interference, improve resolution of phase axis.(8) Curvelet transform has the characteristics of multiscale anisotropic, overcome the wavelet transform the inherent defects in terms of directional characteristics of micro-seismic data processing, such as the edge, Can keep the details of micro-seismic signals suppress random noise interference, while achieving better de-noising effect.(9) Because of micro-seismic signal has its particularity and complexity, a single microscopic micro-seismic signal de-noising its effect, indicating that a single method has its limitations. Combined with a variety of methods, it is explore direction of micro-seismic signal de-noising.