Vertical seismic profiling (VSP) techniques provide a method to measure accurately the seismic velocity and lithologic structure near the borehole. The analysis of a VSP survey can also provide insight into seismic‐wave propagation especially when related to sonic measurements. But VSP and sonic log velocities (or traveltimes) are often found to disagree. Recent field evidence of these differences suggests that the VSP traveltimes are delayed with respect to the integrated sonic times, especially in the deep section (>3000 ft), by about 2.0 ms/1000 ft on the average. The VSP has numerous applications in exploration geophysics, such as calibrating the sonic log. It is thus important to understand why the two measurements differ. Differences in the geometries, source frequencies, and instrumental errors of the two surveys are reviewed. More detailed analysis of seismic wave propagation in the VSP shows that short‐path multiples and velocity dispersion can have a significant delaying effect on the seismic traveltimes. One‐dimensional, wide‐band VSP synthetic seismograms are generated in the frequency domain to study these effects. Different parameters (bandwidth, signal‐to‐noise, layer thickness, multiples, attenuation, dispersion) are varied in the synthetic seismograms. A comparative display of synthetic VSP traveltime minus the integrated sonic time is used to view the effects of these parameters on the synthetic traces. Reasonable variation in noise, layer thickness, bandwidth, and picking method have a small effect on traveltimes. Field data from the Anadarko Basin (4 wells) and an East Texas well are examined with the same technique. From the modeling and field examples, it is found that short‐path multiples can cause a seismic pulse delay of up to 2.0 ms/1000 ft with respect to the integrated sonic log in highly cyclically stratified sections. Velocity dispersion associated with attenuation can have a larger effect, causing up to 7.0 ms/1000 ft delay of the VSP traveltimes with respect to the integrated sonic. These wave propagation effects can explain the observed discrepancy between VSP and integrated sonic times in the deep section.

Reflection and refraction of plane acoustic waves are studied for the case where the sediment is modeled as a porous viscoelastic medium. The model is based on the classical work of Biot which predicts that three different kinds of attenuating body waves may propagate in the sediment. As a consequence when homogeneous plane waves in water are incident to a water-sediment interface, three nonhomogeneous waves are generated in the sediment. In these waves the direction of phase propagation and the direction of maximum attenuation are not the same and particle motion follows an elliptic path. Moreover, the velocity and attenuation of the refracted waves become dependent on the angle of incidence and no “critical” angle occurs. Numerical examples show that the reflectivity of a porous viscoelastic model differs significantly from the case where the sediment is modeled as a viscoelastic solid with constant complex modulus. Finally, because of the frequency dependence of reflectivity in the porous model, it is found to act as a filter with respect to broad band energy.

Computer‐processed multichannel seismic reflection profiles from the crestal zone of the East Pacific Rise near the Siqueiros Fracture Zone have revealed distinct layering in the structure not previously observed in single‐channel reflection records. After processing, which included common depth point gathering, normal move‐out and stack, time‐varying predictive deconvolution, and wave equation migration, three distinct crustal layers emerge as coherent events across the rise crest. The base of the third layer appears to be at a depth of about 2 km below the sea floor and may represent the top of a low‐velocity magma chamber, as postulated by Orcutt et al. (1975, 1976) from the results of ocean bottom seismometer refraction data. This layer appears to thicken away from the rise crest toward the edge of the raised axial block that characterizes the East Pacific Rise (Anderson and Noltimier, 1973). Illustrated in this article is the ability of modern processing techniques as applied to multichannel seismic data to enhance weak arrivals, to minimize sound source bubble‐pulse reverberations, and to remove the diffraction patterns caused by rough topography.

No abstract

The phenomenon of geopressure is essentially stratigraphic in nature. In most cases, its occurrence correlates strikingly well with some mappable geologic characteristics, such as lithology changes, sediment deformation, and faulting. High‐precision velocity estimates can be made from the apparent amplitude variations with offset (AVO) that result from moveout errors, even if the seismic data itself lacks any intrinsic AVO. These velocity estimates provide us with an opportunity to estimate cross‐sections and 3‐D volumes of the gradient of pore pressure with depth from surface seismic data. These cross‐sections and volumes may be obtained through the estimation of seismic interval velocities as a function of depth, subtraction of the shale compaction trend, and the calibration of trend deviations in terms of pore‐pressure gradients. When viewed in combination with stacked seismic sections, the pore‐pressure gradient sections provide the interpreter added information about the hydrogeology of the sediment. In this paper, we show examples of pressure gradients caused by a lithology change, sealing faults, and fluid migration flows. Pressure gradient cross‐sections are also extremely useful for the design of mud densities and casing prior to spudding a well.