Velocity Model Calibration Research Articles

Velocity model calibration for surface microseismic monitoring based on a 3D gently inclined layered equivalent model

Abstract Shale gas has become a major source of natural gas production and has received worldwide attention. Hydraulic fracturing is widely performed to stimulate oil and gas wells with considerable success. Given high-precision microseismic (MS) event locations, we can predict the development trend and region of fracturing and evaluate the stimulation effect, thereby providing technical guidance for subsequent exploitation. An accurate velocity model is essential for MS event positioning. However, simple velocity models, such as the uniform or vertical transverse isotropy (VTI) velocity model, are generally applied to calibrate the velocity model. Despite calibration, the VTI model may still face challenges in obtaining accurate MS event locations. Based on the structural characteristics of shale, we propose a novel local velocity model calibration algorithm for surface MS monitoring. To calibrate the velocity model, the actual strata interfaces are replaced with 3D gently inclined planes. We use very fast simulated annealing to concurrently tune the velocity, depth, and angle parameters of the model. Through the assessment of both the stacked amplitude at the position of the perforation shot and the relocation error of the perforation shot, we determine the ideal velocity model. To evaluate the effectiveness of our approach, we conduct experiments on both a synthetic model and a field dataset, and statistically analyze the location error. The results show that the proposed method obviously reduces the perforation shot relocation error and is well-suited for calibrating velocity models that are close to slightly inhomogeneous layered media.

A physics‐guided neural network‐based approach to velocity model calibration for microseismic data

ABSTRACTThe physics‐guided neural network framework combines the effectiveness of data‐driven and physics‐based models, and it is, therefore, becoming increasingly popular in geophysical applications. We present a physics‐guided neural network–based approach to calibrate velocity models for microseismic data. In our implementation, the physics‐guided neural network comprises of a user‐selected number of fully connected layers, a scaling and shifting layer and a forward modelling operator layer. We input the observed P‐ and S‐wave arrival times to the neural network. In the forward pass, the network's output layer produces normalized P‐ and S‐wave velocities for the subsurface model. The scaling and shifting layer converts the normalized output to realistic velocity values. The forward modelling operator (i.e. a ray‐shooting algorithm) layer computes traveltimes using the velocities from the preceding scaling and shifting layer and the known source–receiver locations. We then evaluate a loss function that compares the predicted traveltimes with the input observed arrival times, and update network's weights and bias parameters. We also use a weight‐based saliency measure to evaluate whether the selected network architecture (i.e. number of hidden layers and neurons) is optimal for the model calibration problem. Finally, using synthetic data examples, we demonstrate that our unsupervised physics‐guided neural network–based approach can provide robust velocity model and uncertainty estimates.

Errors in microseismic events locations introduced by neglecting anisotropy during velocity model calibration in downhole monitoring

Microseismic monitoring is the only tool to provide locations of microseismic events induced e.g. during hydraulic fracturing of unconventional reservoirs. Obtained locations become a baseline for the evaluation of stimulation effectiveness. However, due to several factors, they can be different from the true ones and lead to misinterpretation. Most important is the inaccuracy of the velocity model used in the location procedure, which is often amplified by seismic anisotropy. Even though the anisotropy is a common feature of organic-rich mudrocks, it is often neglected in the current industry practice. The usual approach is to produce a set of isotropic models, valid for the limited subsurface area only, which compensate for anisotropy effects. Here we explore how the accuracy of microseismic events' locations is affected when neglecting the anisotropy during velocity model building using traveltimes of calibration shots. We analyse events' mislocations using the synthetic modelling that simulates an ongoing multi-stage treatment in a vertically-transverse isotropic (VTI) media, monitored with a single vertical borehole. Finally, we provide a quantitative interpretation of location errors expected to be present when VTI anisotropy is being neglected. The most important result is the demonstration of how spatially limited are single-stage isotropic models in terms of correct event locations, despite performing an accurate calibration. Also, we show the limited gain in terms of location accuracy when multi-stage isotropic velocity models are constructed. As the ultimate solution to reduce location errors, we propose to account for anisotropy by inverting VTI anisotropy model parameters even when only part of calibration data is available.

Enhanced surface seismic data processing using simultaneous acquired DAS-VSP data

Distributed Acoustic Sensor (DAS) is increasingly recognized as a viable alternative to downhole geophone arrays for the acquisition of borehole seismic (VSP – Vertical Seismic Profile) or DAS-VSP data. When acquiring surface seismic data in the field, simultaneously acquired DAS-VSP data in an available well or wells can provide full well VSP imaging and significantly enhance surface seismic data processing. Borehole seismic data can provide higher resolution structural images around the borehole as the receivers are located inside a borehole and near the reservoir. In addition, the formation velocity, deconvolution operator, absorption, attenuation (Q value), anisotropy parameters and other reservoir formation information in a certain range around the wellbore can be obtained directly from the VSP data. These can be used to significantly enhance the surface 3D seismic data processing, such as velocity model calibration and modification, static correction, deconvolution, demultiple processing, high frequency restoration, anisotropic migration, and Q-compensation or Q-migration. In this project, Q-migration was conducted with the enhanced Q-field data volume obtained from the joint inversion of both the near surface 3D Q-field data volume from uphole data and the mid-deep layer Q-field data volume from VSP data. The Q-migration results show much sharper and more focused faults and fracture zones. A new 3D formation top structural model was built with much finer details based on the Q-migration data, and three new borehole locations were recommended to the client. The final drilling results on these three new boreholes are very successful with high volume of commercial oil flow.

Velocity calibration for microseismic event location using surface data

Because surface-based monitoring of hydraulic fracturing is not restricted by borehole geometry or the difficulties in maintaining subsurface equipment, it is becoming an increasingly common part of microseismic monitoring. The ability to determine an accurate velocity model for the monitored area directly affects the accuracy of microseismic event locations. However, velocity model calibration for location with surface instruments is difficult for several reasons: well log measurements are often inaccurate or incomplete, yielding intractable models; origin times of perforation shots are not always accurate; and the non-uniqueness of velocity models obtained by inversion becomes especially problematic when only perforation shots are used. In this paper, we propose a new approach to overcome these limitations. We establish an initial velocity model from well logging data, and then use the root mean square (RMS) error of double-difference arrival times as a proxy measure for the misfit between the well log velocity model and the true velocity structure of the medium. Double-difference RMS errors are reduced by using a very fast simulated annealing for model perturbance, and a sample set of double-difference RMS errors is then selected to determine an empirical threshold. This threshold value is set near the minimum RMS of the selected samples, and an appropriate number of travel times within the threshold range are chosen. The corresponding velocity models are then used to relocate the perforation-shot. We use the velocity model with the smallest relative location errors as the basis for microseismic location. Numerical analysis with exact input velocity models shows that although large differences exist between the calculated and true velocity models, perforation shots can still be located to their actual positions with the proposed technique; the location inaccuracy of the perforation is <2 m. Further tests on field data demonstrate the validity of this technique.

The differences in the detectability of perforation shots and microseismic events in the surface monitoring: the attenuation effect

Detection of the seismic signal from perforations or string shots is crucial for calibration of velocity model for surface or downhole monitoring. While in the borehole recordings perforation shots have higher signal to noise ratio of the P-waves than microseismic events, in the surface monitoring recordings it is often opposite. We investigated amplitude spectra of microseismic events and perforation shots in the downhole data and found that amplitude ratios in the band of 20-30 Hz relative to band 100-200 Hz are stronger for the microseismic events. Thus we suggest that detectability of the perforations in surface recorded data is limited by attenuation of the higher frequency signal which increases exponentially with frequency. This suggests that longer duration of calibration shots can improve detectability of string shots at the surface. ARTICLE INFO