Studies On The Downhole Microseismic Monitoring Methods

Microseismic monitoring has been an approach for helping unconventional oil and gas production.The signal produced by hydraulic fracture is recorded by multiple receivers deployed in the monitoring wells or on the ground.We locate these microseismic events according to the waveform information obtained by receivers.The distribution of the microseismic events induced by hydraulic fracturing is used to indicate the geometry of the fracture growth and estimate the stimulated reservoir volume.Therefore,it is important to infer accurate locations of microseismic events.For deep shale fracking,a downhole array in a single well is the most common monitoring system for monitoring microseismic.The downhole monitoring has certain pro and con:such as poor angle converge,a limited number of receivers,and better signal than surface monitoring.A 1D layered model is generally employed in the location procedure because the microseismic events are not too far away from the monitoring well.Our study focuses on the microseismic event location methods,simultaneous update velocity model,and event locations.The methods presented in this study are summarized as following:1.Relative event location method for microseismic eventThe double-difference(DD)location method has long been applied for locating a cluster of earthquakes with data recorded at surface seismic stations.This method has also been used for locating microseismic events with multiple monitoring wells during hydraulic fracturing.Through theoretical analysis,the DD algorithm contains both absolute and relative location information.However,the coefficient of the absolute location is much smaller in magnitude than that of the relative location.Thus the absolute results may rely on the barycenter of the initial location estimates if there are picking errors.We develop a cross double-difference(CDD)approach by using the cross traveltime difference between the P-wave arrival of one event and the S-wave arrival of another event for inversion instead of the arrival time differences of the same phases as in the DD method.The CDD algorithm contains more information on absolute locations than the DD method.Thus,the accuracy of the absolute location is improved while preserving the relative location accuracy.Both synthetic and field data tests indicate that the CDD method can improve the accuracy of both relative and absolute event locations in microseismic clusters.2.Migration-based microseismic location methodsThe migration-based source location methods have been proposed due to their robustness and automatism for detecting and locating weak microseismic events.It also has high computational efficiency by stacking the amplitudes of seismic traces along diffraction traveltime curves.The polarity of the first motion of a signal from the microseismic event is very important for stacking result.The convolutional neural network(CNN)is applied to predict the polarity of the waveform.We utilize the synthetic and field data waveform with known polarity to train the neural network.We expect that the CNN extracts the relationship between the waveform feature and the polarity,and predicts the polarity for the other waveform.Results indicate that the CNN has the strong ability to predict the polarity of the waveform.The traditional migration method needs searching the origin time,then the source location is obtained from the maximum of the image section for the time window under consideration.In order to eliminate the origin time,the imaging function is replaced with the S-P phase arrival time difference and the P-and S-wave arrival differences of receiver pairs,which are calculated by the autocorrelation and cross-correlation,respectively.Results indicate that this method has a lower location accuracy compared with the traditional migration method when the signal-noise-ratio is poor.3.Simultaneous invert event locations and velocity modelDetecting the accurate locations of microseismic events relies on an accurate velocity model.The ID layered velocity model is generally obtained by model calibration from inverting perforation data.However,perforation shots may only illuminate the layers between the perforation shots and the recording receivers with limited raypath coverage in a downhole monitoring problem.Some of the microseismic events may occur outside of the depth range of these layers.The simultaneous inversion method is applied to derive an accurate velocity model covering all of the microseismic events and locating events at the same time.Because of the trade-off between the event locations and velocity model and the poor angle converge,the simultaneous inversion result is unstable.Thus,we apply the CDD method for the simultaneous inversion of a velocity model and event locations using both perforation shots and microseismic data.The microseismic data is used to update both velocities and locations,while the perforation shots are used to update velocities.Moreover,the CDD method could provide accurate locations in both the relative and absolute sense.4.Research on multiple parameters inversionFor geophysical inverse problems,the velocities are strongly coupled with the event locations.Because responses of data to different parameters are different,the elements of the coefficient matrix have different orders of magnitude,which generally lead to ill-conditioned.In this study,we evaluate several strategies to update the 1D velocity model and estimate event locations in downhole microseismic monitoring by conducting numerical experiments.The inversion strategies include:(1)the simultaneous inversion of event locations and velocity model;(2)searching event locations and velocity model alternately;and(3)the joint Neighbourhood Algorithm and grid search method.The coupling between the event locations and velocity model and the limited distribution of the receivers may lead to an ill-conditioned problem for simultaneous inversion.Thus,good initial estimates are essential to produce a stable solution.The alternating inversion decouples event locations from the velocity model,which is less affected by the starting velocity model,but it also results in an approximate solution.Only the joint Neighbourhood Algorithm and grid search method will ensure a stable global solution,but may require substantial computation.Considering the above testing results,we present an effective workflow to update both 1D velocity model and microseismic event location by combining strategy(1)and strategy(2).We first acquire an initial solution with the alternating inversion,and then conduct the simultaneous inversion with the above initial solution.We demonstrate the effectiveness of the workflow by applying to synthetic and field data.