Harry, Dennis L. (Colorado State University) | Sanford, William E. (Colorado State University) | Gehman, Carter (Hess Corp.) | Woodworth, Joshua (Shell Oil Co.) | Damiata, Brian N. (University of California at Los Angeles)
Time-lapse gravity surveys involve gravity campaigns conducted at different times. Differences in gravity measured at given locations between the surveys, after corrections for barometric pressure changes, Earth Tides, and instrument drift, reflect changes in subsurface mass. In this study, we associated these mass changes to changes in groundwater storage. We demonstrate that, in unconfined surficial aquifers, water table elevation changes of ca. 70 cm can be detected, and, if drawdown data are available, specific yield, storativity, and specific storage can be estimated (although specific storage is poorly constrained). We present two case studies, one involving an artificial recharge project in which water table changes are mapped via time-lapse gravity over a ca. 10 km
We examine the application of the transient CSEM method in shallow water (less than 500 m deep) using towed-streamer EM data obtained in a 2009 survey over the Peon gas field, Norway, and consider the in-line electric field component. We compare the results obtained with two different source signatures: a transient square wave and a transient pseudo-random binary sequence (PRBS).
The theory of super-virtual refraction interferometry was recently developed to enhance the signal-to-noise ratio (SNR) of far-offset traces in refraction surveys. This enhancement of SNR is proportional to
This is an update on a project initiated with a grant from SEG''s
Modeling of time shifts associated with time-lapse (4D) seismic surveys is helpful in evaluating reservoir depressurization and inverting for subsurface stress. Here, we discuss time shifts estimated from synthetic seismic data and analyze their dependence on reflector depth and pressure drop inside the reservoir. Accurate time shift measurements between baseline and monitor surveys are obtained by advanced processing techniques that are potentially applicable to field data. Coupled geomechanical and seismic modeling is used to study time shifts for parameters close to those for West Texas petroleum reservoirs. Time-shift leads and lags for P-, PS-, and S-waves are estimated for a wide range (10% to 50%) of effective depressurizations. For the largest pressure drop, P-wave time lags reach 45 ms for reflectors above the reservoir, while S-wave leads reach 90 ms for reflectors below the reservoir. We also investigate the contributions of the deviatoric and volumetric strains to the time shifts, and show that the deviatoric strain is largely responsible for P-wave anisotropy near the reservoir. Time shifts for S-waves and, to a lesser extent, PS-waves, are strongly influenced by the volumetric strain inside the reservoir. Moderate tilt of a rectangular reservoir, or its replacement with an elliptically shaped reservoir of the same aspect ratio, has little influence on time shifts. Potentially, our methodology can be applied to invert for compaction-induced stress fields using simple compartmentalized reservoir models.
A new approach of local time-frequency decomposition is presented, which has high resolution in both time and frequency domain and can represent a signal more sparsely. Seismic data usually belongs to the class of nonstationary signal, so it''s more logical to process the data in time-frequency domain. It''s all known that windowed Fourier transforms (WFT) and wavelet transforms (WT) are two important classes of local time-frequency decompositions, but they can''t have a set of basis whose energy is highly localized in both time and frequency, because of the Heisenberg uncertainty principle. The method introduced here conquers this shortcoming. It is deduced under the framework of inverse problem theory and model parameter estimation methods, with the help of windowed Fourier transforms. The numerical examples demonstrate its advantages and effectiveness.
We present a three-dimensional (3D) model-based inversion algorithm for inverting electromagnetic data. In our approach, the models are described by points in the 3D space and the so-called radial basis functions are used as the interpolation functions for connecting these points. The use of the radial basis functions renders the surface of the target intrinsically smooth. A
A velocity anomaly zone with high contrasts and small wavelength variations is often a challenge for conventional depth imaging. It requires model details that are usually smoothed out because of the noise present in the data. In the eastern Gulf of Mexico, carbonate karst zones are often found over the Florida Escarpment as shown in Figure 1. The karst zones have significantly slower velocities than the surrounding sediments. The size of these karsts is usually a few meters to a few hundred meters. The resulting seismic images below the karst zones are often poor and non-geological (Figure 2). A synthetic modeling study has been conducted to give more insight into the difficulty of running conventional tomography, and to provide some guidelines for modeling the karst zone velocity anomaly. A new velocity model building technique has been developed that combines karst modeling and horizon-driven tomography. The geological information and interpretation are used as input as well as constraints to seismic tomography. The integration of geological interpretation and seismic tomography enable us to derive a detailed velocity model and improve the seismic image in the area.
A low-frequency resistivity survey using ERA electrical profiling (middle gradient array) has been completed over a nickel prospect in Porcupine District, Ontario, Canada. The results of this EM-Resistivity were used to calculate the apparent resistivity and were further inverted using UBC-GIF 3D Frequency Domain Inversion Software, EH3DInv. The results of the inversion have shown the anomalous distribution of electrical properties, consistent with a known geological contact between metavolcanic rocks and an ultramafic pluton. The conductive zone has been interpreted to be located in the interval between 70 and 120 meters. The interpreted conductive zone was further subjected for a drilling program. The drilling program was carried out in March–April 2011 and revealed a 21m thick zone with disseminated to massive mineralized intervals starting at 82 meters, which is consistent with the results of 3D inversion of ERA data.
The increasing need for continuous reservoir monitoring is one of the primary concerns to the oil industry to improve the hydrocarbon recovery factor and production efficiency. Several monitoring scenarios with geophysical methods can be derived including surface and borehole-based methods and their combinations. One is a surface electric current dipole and a vertical electric borehole receiver which has the strongest coupling in detecting the water flood front changes and is easy to implement. The surface-to-borehole electromagnetic if combined with seismic can give excellent resolving capabilities. A modeling study was performed to generate several results based on the given model. This is to support feasibility studies as well as to determine survey acquisition parameters. A 3-layer model was used with a hydrocarbon reservoir in the second layer. The optimum transmitter offset was determined by the modeling result and the value was used for the rest of the experiment. The resistivity of the hydrocarbon reservoir was also varied to observe the received vertical electric field. A time lapse study is relevant for the reservoir monitoring. We built and simulated 3-D model to apply this technology to real reservoirs. In combinations with reservoir simulator results it predicts the outcome of potential surveys. The model is then translated to time lapse fluid changes in order to design the survey layout such that we can get a maximum response.