Luyi Wesley Shen

Nuclear magnetic resonance (NMR) is widely used to characterize fluids in rock pore spaces, but traditional methods have difficulty distinguishing fractures from matrix pores in complex carbonate formations. To address this, we developed a calibration method that integrates X‐ray computed tomography (CT) imaging with NMR to identify fracture‐related T₂ signals. The method quantitatively calibrates NMR T₂ spectra to fracture aperture sizes, improving the accuracy of fracture characterization. Fully saturated fractured samples were used, and fracture fluids were progressively removed using gas displacement techniques. NMR spectra were recorded before and after fluid removal to isolate fracture‐specific signals. Fracture size distributions were estimated from CT images by pixel counting, and porosity was determined by fluid saturation measurements, with corrections for matrix porosity not captured by CT resolution. This workflow extracts fracture distributions from T₂ spectra and establishes a correlation between pore radius (r) and T₂, enabling subsequent applications in core analysis and NMR logging. The method improves differentiation between fractures and matrix pores, enhances the interpretation of NMR data, and can be adapted to heterogeneous reservoir systems.

XiangGuoSi reservoir is a depleted gas reservoir that has recently (in 2014) been converted to an underground gas storage facility. It stores gas in the reservoir during the summer season and produces gas in the winter season. In this work, we present a case report on using InSAR to monitor the mountainous area beneath where the XiangGuoSi gas reservoir is located, along with its supporting pipeline infrastructures. Data, containing 159 scenes, from C-band Synthetic Aperture Radar (SAR) aboard Sentinel-1 satellite is used here, the processing period covered a timespan of 5.6 years. Importantly, we find that the surface deformation is not correlated with the reservoir’s gas injection/extraction history. This indicates that the gas storage’s underground operation does not increase geohazard risk in the area. Further, this indicates the reservoir rock’s pore structure is rather stable even during the cycles of injection/extraction, suggesting a stable reservoir performance even into the far future. Nevertheless, the natural movement of the mountain still poses a landslide risk for the pipeline structure. Given our observed deformation is mostly monotonically downward (subsidence) and in many points, linear, we develop a protocol using 1. the local maximum deformation rate point’s proximity to the pipeline and 2. the rate and total deformation magnitude reported during the monitoring period. After all, this report shows the capability of InSAR as a tool for mapping geohazards for mountainous areas where critical infrastructures are in the vicinity.

The precise relocation of seismic events is critical for many engineering projects. Swarms of minor or micro earthquakes typically reveal stress concentration and spots of greater seismic hazards. Particularly in the context of deep underground mining, advanced techniques that can accurately relocate microseismicities are urgently in demand. Here, we developed a neural network-based modeling training method that can precisely relocate seismicities and invert for velocities at the same time, with preconfigured receiver network locations. Our model can be iteratively improved with field recorded data. We showed that, with roughly eight iterations, we can reasonably resolve for the earthquake locations for both clusters of events, namely spatially distributed with linear pattern or randomly scattered. Our initially trained model, which only focused on events that had a linear distribution pattern, was used as the base for the training of the subsequent models which could better resolve for randomly scattered event locations. Although we stopped at the eighth iteration, the process reported here can be continued, as the model will have a better performance with more iterations.

Dynamic material constants obtained by wave-based methods are different from their static counterparts. Constraining rock's elastic constants’ dynamic-to-static ratios (R_ij) are important for understanding the geomechanical properties of earth's materials, particularly in the context of hydraulic fracturing that requires the knowledge of shale's static elastic constants. Conducting experiments with dynamic and elastic constants’ anisotropy, on top of their pressure dependency, properly accounted for is challenging. Here, we measure suites of dynamic and static elastic constants, with anisotropy fully accounted for, on the shale samples extracted from the Duvernay unconventional reservoir; a comprehensive set of geochemical/petrophysical measurements are obtained too. We observe that the dynamic-to-static ratios are generally not sensitive to the increasing pressures at σ > 50 MPa; we do not find a correlation with the samples’ mineral contents either. However, we find that R_ij strongly correlates to the dynamic elastic constants except for the R₁₁. The correlation between R_ij, particularly R_i3, and the dynamic elastic constants can be explained by the sedimentary rocks’ compactness and the horizontal void spaces parallel to the rock's laminated bedding planes.

The ground surface deformation induced by shield tunnels passing through enclosure structures of existing tunnels is a particular underground construction scenario that has been encountered in Wuhan Metro Line 12 engineering cases in China. Timely ground deformation prediction is important to keep shield tunneling safe. However, the classic ground deformation theory is difficult to accurately predict for this ground deformation. This paper develops a semi-analytical method to predict ground heave considering the space effect in this engineering condition. Based on the improved ground deformation theory, a novel deformation prediction method for the ground and enclosure structure is derived and combined with Kirchhoff plate theory. Comparing with field deformation measurements, the maximum difference between the measured and calculated deformation is 14.6%, which demonstrates that the proposed method can be used to predict the ground heave induced by shield tunnels passing through the enclosure structure of existing tunnels. The parameters of the underground diaphragm wall used in Wuhan Metro Line 12 are further studied in detail. The results show that the ground heaves have a positive correlation with the embedded ratio of the diaphragm wall, but a negative correlation with its elastic modulus and thickness. However, the thickness and embedded ratio have a limited effect on ground heaves. This study provides a technical reference for optimizing the setting of enclosure structures in order to protect existing buildings.

The ground surface deformation induced by shield tunnels passing through enclosure structure of existing tunnels is a particular underground construction scenario, which is encountered in Wuhan metro line 12 engineering cases in China. The classic ground deformation theory is difficult to accurately predict this ground deformation. This paper develops a semi-analytical method to predict ground heave considering space effect in this engineering condition. Based on improved ground deformation theory, a novel deformation prediction method of ground and enclosure structure is derived combined with Kirchhoff plate theory. Comparing with field deformation measurements, the maximum difference between measured and calculated deformation is 14.6%, which demonstrating that the proposed method can be used to predict the ground heave induced by shield tunnels passing through the enclosure structure of existing tunnels. The parameters of underground diaphragm wall used in Wuhan metro line 12 are further studied in detail. The results show that the ground heaves have positive correlation with embedded ratio of diaphragm wall, but negative correlation with its elastic modulus and thickness. But the thickness and embedded ratio has a limited effect on ground heaves. This study provides a technical reference for optimization setting of enclosure structure in protecting existing building.

Summary: Successful examples of hydraulic fracturing (HF) have led to a general consensus that fractures enhance hydrocarbon production as it connects isolated pockets filled with oil. However, the fracture’s impact on the hydrocarbon recovery from rock’s pores, which can account for a significant amount, is not well studied but experimentally investigated here. Uniquely, we fill our samples' fractures with a slime made of guar gum and heavy water (D2O). Such slime cannot penetrate into pores, and the heavy water does not generate nuclear magnetic resonance (NMR) signals detectable by our apparatus. Using such slime as a fracture filling material allows us to isolate and eliminate the NMR signals from fractures. Subsequent huff ‘n’ puff (HnP) experiments show that fracture results in a decline of the total HnP recoverable oil. We reasoned that fracture reduces the treatment gas’ sweeping efficiency. This issue can be partially mitigated by injecting N2 instead of CO2; N2 can enter rock’s pores more efficiently as a treatment agent. Nevertheless, N2 HnP still suffers a noticeable reduction in the total recovery for samples with smaller pores. Our experiments provide an important new insight into the development of unconventional hydrocarbon resources. Maximizing fracture intensity in the field HF operation may negatively impact the later HnP oil recovery. At last, this observation is only possible with our unique experiment design; a misleading and opposite finding will be reached if one directly compares the results from before and after fracturing the samples. We reinforce that pore structures are changed during the fracturing process in our experiment (and possibly other researchers’ work); such change needs to be properly accounted for to assess fracture’s impact on pore fluid movement fairly.

Fault slip is controlled by the normal and shear tractions on a fault plane. A full understanding of the factors influencing induced seismicity requires quantitative knowledge of the in situ stress tensor and fluid pressure. We analyze these variables for a 200 × 200 km region with active hydraulic fracturing near the city of Red Deer, Canada. The levels of induced seismicity in the area were generally low before March 4, 2019, when local residents felt the M_W 3.8/M_L 4.2 event. We use geophysical logs and pressure tests within the targeted Duvernay Formation to construct maps of ambient pore pressure, vertical and minimum horizontal stresses. The maximum horizontal stress is constrained by the focal mechanism inversion and borehole‐based estimation method. We find that a broad range of orientations are susceptible to slip and small perturbations of fluid pressure would promote displacement. This suggests that the differential variations in pore fluid pressure in the target formation may provide a metric of slip susceptibility; a map for the study area is developed. Areas of high susceptibility correlate with those experiencing higher levels of induced seismicity, except for the Willesden Green oil field that has similarly elevated susceptibility and active hydraulic fracturing operations. The methods and results demonstrate how more quantitatively constrained in situ stresses developed from an ensemble of real field measurements can assist in assessing fault stability and in developing metrics for slip susceptibility.