A shallow borehole array was employed to record microseismic waveforms during multi-stage hydraulic fracturing operations near Fox Creek, Alberta. A total of 530 events with magnitude larger than MW 0.2 were used to calculate moment tensors, based on the 3C amplitudes and polarities of the direct P-wave. By classifying the best double-couple solutions of the retrieved moment tensors, three groups of distinct focal mechanisms were identified, of which two have predominantly double-couple source mechanisms. The moment tensors of group three exhibit significant non-double-couple components, which needs further interpretation. Based on the resulting source mechanisms, orientations of three principal stresses were estimated. Observed linear alignments within the spatial distribution of microseismic events aided in identifying the fault plane. Presentation Date: Thursday, October 18, 2018 Start Time: 8:30:00 AM Location: 210C (Anaheim Convention Center) Presentation Type: Oral
Abstract We present a novel physics-guided neural network to estimate shear-tensile focal mechanisms for microearthquakes using displacement amplitudes of direct P waves. Compared with conventional data-driven fully connected (FC) neural networks, our physics-guided neural network is implemented in an unsupervised fashion and avoids the use of training data, which may be incomplete or unavailable. We incorporate three FC layers and a scaling and shifting layer to estimate shear-tensile focal mechanisms for multiple events. Then, a forward-modeling layer, which generates synthetic amplitude data based on the source mechanisms emerging from the previous layer, is added. The neural network weights are iteratively updated to minimize the mean squared error between observed and modeled normalized P-wave amplitudes. We apply this machine-learning approach to a set of 530 induced events recorded during hydraulic-fracture simulation of Duvernay Shale west of Fox Creek, Alberta, yielding results that are consistent with previously reported source mechanisms for the same dataset. A distinct cluster characterized by more complex mechanisms exhibits relatively large Kagan angles (5°–25°) compared with the previously reported best double-couple solutions, mainly due to model simplification of the shear-tensile focal mechanism. Uncertainty tests demonstrate the robustness of the inversion results and high tolerance of our neural network to errors in event locations, the velocity model, and P-wave amplitudes. Compared with a single-event grid-search algorithm to estimate shear-tensile focal mechanisms, the proposed neural network approach exhibits significantly higher computational efficiency.
Abstract The subduction model of the Neo-Tethys during the Early Cretaceous has always been a controversial topic, and the scarcity of Early Cretaceous magmatic rocks in the southern part of the Gangdese batholith is the main cause of this debate. To address this issue, this article presents new zircon U–Pb chronology, zircon Hf isotope, whole-rock geochemistry and Sr–Nd isotope data for the Early Cretaceous quartz diorite dykes with adakite affinity in Liuqiong, Gongga. Zircon U–Pb dating of three samples yielded ages of c. 141–137 Ma, indicating that the Liuqiong quartz diorite was emplaced in the Early Cretaceous. The whole-rock geochemical analysis shows that the Liuqiong quartz diorite is enriched in large-ion lithophile elements (LILEs) and light rare-earth elements (LREEs) and is depleted in high-field-strength elements (HFSEs), which are related to slab subduction. Additionally, the Liuqiong quartz diorite has high SiO 2 , Al 2 O 3 and Sr contents, high Sr/Y ratios and low heavy rare-earth element (HREE) and Y contents, which are compatible with typical adakite signatures. The initial 87 Sr/ 86 Sr values of the Liuqiong adakite range from 0.705617 to 0.705853, and the whole-rock ϵ Nd ( t ) values vary between +5.78 and +6.24. The zircon ϵ Hf ( t ) values vary from +11.5 to +16.4. Our results show that the Liuqiong adakite magma was derived from partial melting of the Neo-Tethyan oceanic plate (mid-ocean ridge basalt (MORB) + sediment + fluid), with some degree of subsequent peridotite interaction within the overlying mantle wedge. Combining regional data, we favour the interpretation that the Neo-Tethyan oceanic crust was subducted at a low angle beneath the Gangdese during the Early Cretaceous.
We have developed a novel regularized approach to estimate a composite focal mechanism for microseismic events that share a similar source mechanism. The method operates by minimizing the weighted misfits of the SH/P amplitude ratios (in absolute sense and logarithmic scale) and P-wave polarities, using a regularization parameter determined from the trade-off curve for these values. This approach overcomes the low signal-to-noise ratio (S/N) and single-event azimuthal gaps that may otherwise limit the effectiveness of sparse surface arrays. Compared with focal mechanisms derived from P-wave polarity or amplitude-based methods, our regularized approach reduces the multiplicity of solutions and avoids the use of signed amplitude ratios, which may be ambiguous for data with low S/N. We apply our method to a set of 13 microseismic events recorded during hydraulic-fracture stimulation of the Marcellus Shale in West Virginia and Pennsylvania, USA, yielding a strike-slip focal mechanism accompanied by a minor normal component. Our solution is similar to previously reported focal mechanisms in this area. Jackknife analysis, which tests stability of the inversion based on random sampling of the observation, indicates 95% confidence intervals of 1° and 2°, respectively, for the plunge and azimuth of the P and T axes. By analyzing the event subsets, outliers are identified and the assumption of a single dominant focal mechanism is validated. Numerical modeling demonstrates that our approach is robust in the presence of variations of up to 0°–10° and 0°–35°, respectively, for the plunge and azimuth of P and T axes of the focal mechanisms of these events. Sensitivity analysis using synthetic data also indicates that the algorithm is tolerant to mispicks as well as errors in polarity and amplitude ratio. In the presence of some dissimilar focal mechanisms, the dominant focal mechanism can be reliably estimated if at least 70% of the events have similar source mechanisms.
The Fanshan alunite deposit, located in Cangnan County, Zhejiang Province, is the largest alunite deposit in China, which developed a lithocap with an extensive advanced argillic alteration. Surface alteration mapping, through field geology and shortwave infrared (SWIR) spectroscopy analyses, has defined three alteration zones including Na-alunite-dickite-pyrophyllite-topaz alteration, K-alunite-pyrophyllite-phengitic muscovite alteration, and chlorite-illite alteration from center to margin. Alunite varies from K-rich to Na-rich and has different paragenesis and texture. Na-alunite, with quartz, pyrophyllite, dickite, and topaz, mainly developed in Pingpengling area of Fanshan, forming a typical advanced argillic alteration. The wavelength position of the alunite –OH spectral absorption feature at ∼ 1480 nm (Pos1480) varies from 1478 nm to 1492 nm with increasing content of Na in alunite. There exists a positive correlation between Na/K ratio value and Pos1480 of alunite. The TESCAN Integrated Mineral Analyzer (TIMA), coupled with LA-ICP-MS analyses of alunite has been conducted, aiming to systematically characterize the texture and mineral chemistry of alunite. Fanshan alunite is rich in Ga, V, rare earth elements (REEs), and large ion lithophile elements (LILEs). Na-alunite has higher LILEs and light rare earth elements (LREEs) than K-alunite. Thorium, U, Li, and B, vary greatly in the Na-alunite and K-alunite. The ratio of Na/K in alunite, together with Pos1480 has been used to vector the magmatic-hydrothermal center. Furthermore, the content of Pb and ratios of Rb/Sr*1000, Sr/Pb, La/Pb, have illustrated the consistent results as Na/K and Pos1480. Based on the spatial distribution of these vectors in conjunction with the occurrence of Na-alunite, we concluded that Pingpengling area is the most proximal area to the magmatic-hydrothermal center, which may imply a possible causative intrusion underneath. Further exploration is suggested to be conducted under Pingpengling to verify if there is porphyry Cu(-Au-Mo) mineralization.
SUMMARY Estimating microseismic event locations is important for applications of geophysical monitoring, including hydraulic fracturing and carbon-capture and storage. Field sites for these applications are typically located in sedimentary basins that include finely stratified sediments, particularly around the target depth of the application. The fine stratification causes vertical transverse isotropy (VTI) for seismic wave propagation. In addition, such sediments often exhibit a vertically fractured rock mass that can cause horizontal transverse isotropy (HTI). Therefore, geophysical monitoring can be strongly affected by the occurrence of anisotropy caused by sets of aligned vertical fractures in finely horizontally layered media. While both HTI and VTI theories exist, a more efficient approximation to include both effects is by effective orthorhombic (ORT) models. To account for such anisotropy in microseismic monitoring, we simultaneously estimate ORT parameters, perforation shot locations, and microseismic event locations with Bayesian methods based on direct P-wave arrival times. A comparison to a HTI parametrization is carried out to examine anisotropy-model choice. The quasi-P-wave group velocities in HTI and ORT media are approximated by linearization. Anisotropy parameters are estimated with Markov chain Monte Carlo sampling that includes parallel tempering and principal-component diminishing adaptation to ensure efficient sampling of the parameter space. In contrast to deterministic inversion, our probabilistic non-linear approach includes uncertainty quantification by approximating the posterior probability density with an ensemble of model-parameter sets for effective anisotropy parameters, microseismic event locations, and horizontal locations of perforation shots. The noise standard deviation of P-arrival times is also treated as unknown. The inversion is carried out for simulated data, and for data from a physical laboratory model. In the latter case, an anisotropic layer is represented by a phenolic canvas electric material, and a star-shaped surface-receiver configuration is used to record microseismic signals. Results show that obtaining unbiased event locations requires an appropriate choice of anisotropy model and the ability to resolve anisotropy parameters. The resolution of anisotropy parameters requires significantly more data information from microseismic acquisition than required for isotropic models. Therefore, we study several acquisition scenarios for simulated and laboratory data. Assuming an HTI model in the inversion when data originate from an ORT medium causes systematic errors in event locations. However, appropriate resolution of ORT parameters requires a large acquisition aperture, an accurate perforation-shot timing, and the combination of surface acquisition with a vertical downhole array. These scenarios provide new knowledge about field requirements to produce sufficient information for the resolution of microseismic event locations in the presence of ORT effects in the data.
Accurate microseismic event location requires a thorough knowledge of the velocity model between the treatment well and monitoring well. This velocity model is routinely built by integrating sonic logging data with perforation monitoring data. Conventional approaches of building such a velocity model require that the perforation origin times must be exactly known, so that the travel times of the initial P-waves can be computed and then employed to calibrate the velocities. However, in actual situations, the perforation origin times are very difficult to measure precisely, so in this paper, we propose a new method to invert the velocity model without knowing the perforation origin time. Numerical experiments demonstrate that although the velocity models obtained by our method show some discrepancies, their application introduces unobservable impact on the event location result due to the existence of the picking errors.
A comprehensive dataset from 594 fracturing wells throughout the Duvernay Formation near Fox Creek, Alberta, is collected to quantify the influences of geological, geomechanical, and operational features on the distribution and magnitude of hydraulic fracturing-induced seismicity. An integrated machine learning-based investigation is conducted to systematically evaluate multiple factors that contribute to induced seismicity. Feature importance indicates that a distance to fault, a distance to basement, minimum principal stress, cumulative fluid injection, initial formation pressure, and the number of fracturing stages are among significant model predictors. Our seismicity prediction map matches the observed spatial seismicity, and the prediction model successfully guides the fracturing job size of a new well to reduce seismicity risks. This study can apply to mitigating potential seismicity risks in other seismicity-frequent regions.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
