Second EAGE Workshop on Borehole Geophysics

For large 3D VSP surveys that use large 3-component downhole receiver arrays to record of all the different wavefield components, it is not easy to apply conventional wavefield separation methods such as parametric decomposition or model based rotation. These methods usually fail for very large offsets because they are strongly dependent on the velocity model. In this paper we present a simple but efficient workflow for wavefield separation of large 3D VSPs. The workflow combines a wave-by-wave wavefield separation method with a covariance matrix method. The effectiveness of this technique is confirmed by application to 3D VSP data from Saudi Arabia.

Simultaneous multiwell VSP data have been acquired using fibre optic cables in producing wells as distributed acoustic sensors. The measurement apparatus was retrofitted to the fibre optic cables installed for other purpose with completion of the wells. Data were acquired with no other instrumentation in the well and without disturbing the normal operation of the wells.

The paper discusses the application of VSP multiple analysis technique to guide multiple removal procedures in surface seismic and 3D VSP data processing. The results of the VSP multiple analysis from two wells in the Arabian Gulf helped explain the variability in the quality of the well-to-seismic ties across the OBC survey and optimize de-multiple workflows during the surface seismic data processing.

With the increased number of horizontal wells being used to produce unconventional plays, there is a growing inventory of wells that can benefit from the use of interferometric processing methods. Seismic interferometry for VSPs in horizontal wells can move the surface shots to behave as if they are located at the positions of the receivers in the borehole. It removes the overburden travel path from the redatumed traces without using any velocity information or even surface shot statics. So an image can be made below the borehole using only the local velocity near the borehole. The added benefit of redatuming the shots into the borehole is that the upcoming reflected energy from changes in the lithology below the borehole can act like deep seismic sources allowing imaging of events above the borehole. It is also possible to redatum the receivers to the surface, making each shot into a receiver. This transforms the VSP data set into surface seismic data. I present some practical aspects of the method using synthetic examples to create a more intuitive understanding of the method.

The development of long arrays of; non-permanent, 3-component, high pressure, high temperature down-hole receivers, was escalated in the 1990s by the requirements of Vertical Seismic Profile (VSP) recording. The advent of such arrays increased the feasibility of acquiring micro-seismic surveys. Micro-seismic events are generated in many ways and even small movements in rocks will generate micro-seismic events. Some basic micro-seismic processing is covered before describing some of the many links between VSP and micro-seismic. Many opportunities have been opened up to extend 3D VSP techniques as a result of shared acquisition with micro-seismic monitoring.

Techniques for computation and correction of NMO velocities in VTI media are widely employed. However, there are discrepancies between the anisotropic parameter η values computed based on VSP data and those based on surface seismic data; this prevents the direct use of VSP-derived η values in surface seismic data processing. This paper discusses the causes of these discrepancies and proposes methods for compensating for them using actual full-azimuth seismic data, zero-offset VSP data, and Walkaway VSP data from an oil field in China. Our studies indicate that conventional computation of Vnmo, (based on the horizontal layered media and isotropy assumption, using the conventional NMO formula within a small offset range and relying on the principle of flattening CMP gathers), does not yield correct anisotropic parameters, In practice, surface seismic velocity must be corrected with the velocity computed using zero-offset VSP data. Surface seismic data are then corrected using VTI anisotropic parameters computed based on Walkaway VSP data and accurate seismic imaging results can be achieved.

Inelastic attenuation, quantified by Q, the seismic quality factor, has considerable impact on surface seismic reflection data. A new method for interval Q-factor estimation using near-offset VSP data is based on an objective function minimization measuring the difference between cumulative Q estimates and those calculated through interval Qs. To calculate interval Q, we use all receiver pairs that provide reasonable Q values. To estimate Q between two receiver levels, we use directly the equation that links amplitudes at different levels and can provide more accurate Q values than the spectral ratio method. To improve interval Q-estimates, which rely on travel times, we use a high-accuracy approach in the frequency domain to determine time shifts. Application of this method to real data demonstrates reasonable correspondence between Q estimates and log data.

Shear waves propagating through an anisotropic medium split into two approximately orthogonal phases with different velocities. For a single set of fractures, the faster shear wave is polarized along the fracture planes, while the slower shear wave is polarized in a direction orthogonal to the fracture planes. For multiple sets of fractures, the faster shear is not polarized along one specific fracture plane. In this paper, shear-wave splitting analysis for walkaround VSP data was applied for the characterization of multiple sets fractures. Dispersion of the downgoing S- and reflected P-waves was also investigated as a way to discriminate between open and closed fractures.

Although 3D surface seismic imaging has been the primary tool for geophysical reservoir monitoring to date, vertical seismic profiling (VSP) has characteristics that make this technology an attractive proposition for time-lapse monitoring. However, 3D VSPs are not yet widely applied for time-lapse (4D) monitoring. To better understand the value of this technology as a potential 4D reservoir monitoring tool for a waterflood surveillance project, we conducted a comprehensive feasibility study. Through this study we have been able to compare and contrast 3D/4D VSP data processing capabilities of a few vendors in the industry.

Time-lapse three-dimensional (3D) vertical seismic profiles (VSPs) are an important component of the monitoring, verification, and accounting plan for the Illinois Basin –- Decatur Project. The VSPs will be used to provide information on CO2 plume development, demonstrate containment of the CO2 in the storage formation, and provide data to verify and update models and simulations over the life of the project. VSPs are more economical to acquire and process than surface seismic data and cause less disruption to local landowners as they have smaller acquisition footprints. The 3D VSPs are being acquired with a permanent geophone array. Three 3D VSP surveys have been acquired at the site to date: two baseline surveys and one monitoring survey acquired after ~70,000 tonnes of CO2 had been injected. Baseline 2 and Monitoring Survey 1 data have high repeatability as evidenced by the NRMs repeatability metrics. The final difference displays do not provide conclusive results regarding CO2 movement in the Mt. Simon formation; however, the NRMS depth slice at the injection interval shows higher NRMS values that may be suggestive of the presence of CO2. Future surveys are expected to produce more conclusive results as the volume of injected CO2 increases.