First EAGE/ACGGP Latin American Geophysics Workshop

The theoretical basis for azimuthal AVO analysis is the Zoeppritz equation derived for arbitrarily anisotropic media. Due to the complexity of this equation, it is advantageous to work with a linearization of this, leading to an equation with 9 free parameters.

In this article, we propose an integrated workflow for reconciling multiple 3D seismic surveys with production, petrophysical and core data, while honoring the scales of each measurement. The methodology results in a flow model, which is consistent with each of the measurements. The proposed methodology overcomes the limitations encountered by many existing techniques with regards to vertical resolution, time-to-depth conversion and the relationship between seismic response and reservoir properties. The result of the petrophysical seismic inversion is a fine-scale shared earth model in depth that is consistent with both log and seismic data and can be used for reservoir performance prediction. The execution of the workflow resulted in no pore volume or transmissibility multipliers and no numerical aquifers to achieve a history match. Multiple model scenarios can be tested by altering the underlying a priori geologic stratigraphic framework and/or initial reservoir property assignments. Each scenario will meet the criterion of fitting all the data, seismic, well, and production, at all scales.

Multi-attribute seismic analysis has been widely applied to characterize the large scale fracture corridors (close to seismic resolution) affecting Cretaceous and Jurassic carbonate formations in the Middle East region, where a number of reservoirs comprise tight carbonate lithology with bulk of the production from fractures. In such context, fracture swarms/corridors correspond to seismic and sub-seismic faults which extend to tens or hundreds of meters. The method applied for fracture delineation is based on combined use attributes sensitive to discontinuities. Deliverables are fracture index maps/volumes which are calibrated against fracture characteristics observed at wells. These results can be used further as initialization maps for the Discrete Fracture Network. Multi-attribute analysis appears to be an efficient approach for detecting and characterizing fractured zones, even in absence of azimuthal attributes.

In Rubiales and other Llanos fields in Colombia, conditions are adequate for accurate mapping of lithology based on elastic inversion of seismic data and petrophysical estimation of shale volume fraction. We jointly estimate elastic isotropic medium parameters and mass density from inversion of seismic data, and characterized the sensibility of the parameters to the reservoir lithology, porosity and fluids. By means of various techniques of statistical inference we derived reliable estimation of shale volume fraction and total porosity, calibrated to well-log data. Sensitivity of the seismic data to fluids (heavy oil and brine) is smaller than for lithology at this field. A combination of spatial well information with elastic parameters obtained from the seismic inversion allows for an assertive inference of the spatial distribution of oil-sands. The oil-sand estimator showed positive agreement with the well-log information with more than 70% success in classification of oil-sand samples. The results of this study contributed to an accurate interpretation of the area and a 3D description of the reservoir, with a positive impact in the field development.

Reservoir characterization is a process that describes the rock and fluid properties for a volume of the subsurface that produces or has the potential to produce hydrocarbons. This process integrates measurements into models of the subsurface from the disciplines of geology, geophysics and reservoir engineering, e.g. well log, seismic, and dynamic pressure data, respectively.

A match filter using a synthetic seismogram with an appropriate zero phase wavelet that matches to frequency content within the seismic data can restore any attenuated frequencies in the seismic wavelet back to full amplitude with near perfect "zero phase" alignment. This process which is available as part of the line tying facility of the WinPICS Seismic Workstation provides a simple very effective alternate approach to improving seismic resolution allowing reservoir characterization of reservoirs that were previously beyond the resolution of the conventionally processed seismic data. The match filter process requires final stack traces for operator design but once designed the operator can be applied to gather traces allowing AVO analysis for reservoir characterization where this was not previously possible

In complex land areas with rugged topography, seismic imaging alternatives are usually limited to Kirchhoff prestack time and depth migration, often used in combination with refraction and reflection tomography. This is not the case for marine data, where wave-equation and reverse-time migrations have become the norm; nor is it the case for some land areas in the Middle East, for which high-density, wide-azimuth acquisition geometries allow the use of advanced imaging methods. Why the discrepancy? The reason is partly a difference in data quality (extremely noisy data will possibly defeat attempts at advanced imaging), and partly a difference in acquisition parameters, which themselves are often influenced by anticipated data quality (inadequately sampled data will certainly defeat attempts at advanced imaging). In this paper, we address the problem of spatial sampling for noise attenuation, Kirchhoff and more advanced imaging, and velocity estimation, and we point out the mismatch between the theoretical sampling requirements for advanced imaging and the realities of present-day land acquisition in rugged areas. Perhaps we cannot always achieve a sufficiently well sampled data volume for wave-equation migration by acquisition and seismic data interpolation in rugged areas. Nevertheless, we suggest that moving towards the goal of adequate sampling whenever possible (for example, using denser receiver grids with existing source grid spacing and lower effort per source) will improve our images even when we use standard imaging methods.

An acoustic approach to description of P-wave travel times in anisotropic media is needed in order to avoid the solution of elastic wave equations. Type of stone, lateral and vertical facies variations or preferential direction of fracturing, are some of the causes to produce wave velocity dependence with the angle of propagation in the medium. In this work, we present an equation to the approximation of the relation dispersion for a TTI media. The procedure used in the approximation reduces the fourth order differential equation used to model the P wave to a second order equation. Numerical examples of seismic modeling and migration of TTI media are used to validate the new equation. RTM migration, using the REM expansion in time and pseudoespectral method in space, shows that the P-wave equation improves the migrated image in a VTI media and produces also a stable numerical solutions.

After the synthetic data inversion examples, real 3D data sets have been undertaken by the industry for the last several years. As field data is dominated with P waves, one feasible approach is to use the acoustic approximation. The full waveform inversion (FWI) determines parameters related to the acoustic wave equation but mostly velocities by minimizing the misfit between the observed data and the model data. It has shown tremendous potential especially in 3D wide offset acquisitions. The technique had been used mostly with marine data acquisition. This includes wide azimuth streamer, ocean bottom surveys geometry where the advantage of FWI has convinced the oil industry to pay close attention to the technology and apply it in complex geological settings. In this paper we demonstrate that FWI has its place using land types of shooting geometry in order to derive velocity fields especially for depth imaging .

Understanding the geological-structural environment through regional studies to characterize the velocity field from surface to reservoir has more geological sense. In Sonda de Campeche there are strong velocity variations in the Tertiary rocks that affect the shape and dimensions of structures in time (pull-up or pull-down). The study, modeling and understanding these effects help to propose the best solution and the best image in depth of the reservoir with time-depth conversion process. In this paper, we focus on detailing the interpretation around the fault zones (thrusts and growth faults) of events corresponding to carbonate rocks, to populate interval velocity models with which we obtain the time-depth conversion of the seismic data in the Sonda de Campeche. It was concluded that the seismic data require additional processes to improve the image of events in the shadow zone of faults. Finally, we performed 100 realizations of interval velocity models thus generated a probability distribution of the rock volume. This allows us to represent the uncertainty in the velocity field in areas where there is not well information.