Roy White

The accuracy of estimating Q from seismic data

A major aim of seismic interpretation is the inference of petrophysical properties of reservoir rocks. Because the inversion from seismic to petrophysical characteristics is far from unique, this task requires a range of seismic parameters, prominent among which are seismic velocity, impedance, and Poisson’s ratio. The inclusion of seismic absorption in this list could add desirable complementary information. For example, absorption may be more sensitive to clay content than seismic velocity (Klimento and McCann, 1990). However seismic absorption is difficult to measure, particularly over depth intervals as short as most reservoir intervals.

Properties of instantaneous seismic attributes

A seismic trace can be represented as the real part of a complex‐valued signal. Attributes termed amplitude envelope (or instantaneous amplitude), instantaneous phase, and instantaneous frequency are commonly computed from the complex seismic trace and displayed as colored sections or overlays for interpretational purposes. These attributes can be collectively termed instantaneous attributes since they concisely and quantitatively describe the seismic waveform (or character) at any sample point. As such, they can be extremely useful in correlating seismic events. But there are pitfalls in attaching simple physical meanings to them and their application for interpretational inferences beyond correlation demands careful evaluation based on accurately calibrated synthetic seismograms. For example, instantaneous frequency can validly be negative.

Thin‐bed AVO: Compensating for the effects of NMO on reflectivity sequences

Estimating the amplitude versus offset (AVO) gradient for thin beds is problematic because of offset-dependent tuning and NMO stretch. When AVO analysis is carried out before NMO correction, the nonparallel nature of the NMO hyperbolas results in differential interference conditions at each offset and complicates AVO interpretation. If AVO analysis is carried out after NMO correction, the data bandwidth is distorted and corrections must be made to recover the true AVO response. A correction for NMO stretch can be applied to Fourier spectra obtained after windowing the NMO-corrected prestack data. This approach requires knowledge of the seismic wavelet but seems to be relatively insensitive to noise in the data or uncertainties in the wavelet estimation. The technique allows the interference conditions to be made independent of offset and the correct AVO gradient relative to the normal incidence amplitude to be recovered.

How accurate can a well tie be

There appears to be a body of opinion among geophysicists that well‐log synthetic seismograms and seismic data cannot be expected to match one another very closely. For example, Simmons and Backus (Geophysics, 1996), in a paper on impedance estimation, state “…as is most often true, well‐log‐based synthetic seismograms do not agree well with observed seismic data.” Norman Neidell, in conjecturing on the conditions that generate coherent multiple reflections (TLE, November 1993) comments that explorationists often encounter synthetic seismogram fits that are less than satisfactory. That may be true, but it is not inevitable. We wish to counter these pessimistic views by showing an example of how good a well tie can be.

Well tie for broadband seismic data

ABSTRACT The seismic industry is increasingly acquiring broadband data in order to reap the benefits of extra low‐ and high‐frequency contents. At the low end, as the sharp low‐cut decay gets closer to zero frequency, it becomes harder for a well tie to estimate the low‐frequency response correctly. The fundamental difficulty is that well logs are too short to allow accurate estimation of the long‐period content of the data. Three distinctive techniques, namely parametric constant phase, frequency‐domain least squares with multi‐tapering, and Bayesian time domain with broadband priors, are introduced in this paper to provide a robust solution to the wavelet estimation problem for broadband seismic data. Each of these techniques has a different mathematical foundation that would enable one to explore a wide range of solutions that could be used on a case‐by‐case basis depending on the problem at hand. A case study from the North West Shelf Australia is used to analyse the performance of the proposed techniques. Cross‐validation is proposed as a robust quality control measure for evaluating well‐tie applications. It is observed that when the seismic data are carefully processed, then the constant phase approach would likely offer a good solution. The frequency‐domain method does not assume a constant phase. This flexibility makes it prone to over‐fitting when the phase is approximately constant. Broadband priors for the time‐domain least‐squares method are found to perform well in defining low‐frequency side lobes to the wavelet.

Broad-band Well Tie

By extending the low-frequency content, broad-band seismic data lessens the dependence of seismic inversion on a background model. Despite that, the merging of the inverted seismic with a very low frequency background model poses a new problem: the duration of a well-log synthetic seismogram is often seriously inadequate for defining the sharp spectral decay to zero frequency. To address this issue, a practical approach based on multi-taper spectral analysis is proposed to determine the low-frequency spectral decay of the seismic well tie wavelet. Furthermore the impact on seismic inversion is demonstrated.

Quantitative controls on 3-D seismic processing for prestack attribute analysis

Accurate prestack analysis relies on good‐quality seismic data. The seismic data volume should be organized into gathers of traces that sample the same subsurface region from a range of incident angles, with the change in reflected waveform across these traces being dependent on the rock properties and fluids at the reflecting boundary. Further, the signal to noise (S/N) ratio should be high enough so that the change in waveform and any anomalous behavior are accurately detectable. The aim of the processing scheme is to produce data in such a format without introducing artifacts. The main factors affecting prestack data quality can therefore be broken down into S/N ratio, imaging to common reflection point gathers, lateral resolution, and vertical resolution.

Wavelet Estimation for Broadband Seismic Data

The volumes of broadband seismic data acquired and processed by the industry have grown rapidly. There is also an increasing emphasis on the benefits of broadband seismic for quantitative interpretation. The bottleneck for achieving a satisfactory quantitative interpretation and subsequently reservoir parameter estimation is the well tie, a process through which the seismic wavelet is estimated. However, broadband seismic data pose a challenge for well ties as the duration of the well log is often inadequate to estimate the low frequency decay towards zero frequency. Three distinctive techniques, namely parametric constant phase, frequency domain least-squares with multi-tapering and Bayesian time domain with broadband priors, are introduced in this paper to provide a robust solution to the wavelet estimation problem for broadband seismic data. A case study from North West Shelf Australia is used to analyse the performance the proposed techniques. Generally, when the seismic data is carefully processed then the constant phase approach would likely offer a good solution. Broadband priors for the time domain least-squares method are found to perform well in defining low-frequency side-lobes to the wavelet.

Optimised Spectral Merge of the Background Model in Seismic Inversion

Seismic inversion generates low-frequency modulations if the frequency content of the background model does not merge smoothly into the spectrum of the inverted seismic data. An incorrect low-frequency phase is another source of inversion artefacts. We present a systematic method of selecting a background model and low-frequency phase response that minimises the misfit at the spectral merge. The method compares well-log and seismic relative impedances computed at best-match well-tie locations over a range of background models and low-frequency phase corrections. The background models are specified by a low-cut corner frequency and the phase corrections by the phase intercept at zero frequency. The method is illustrated by application to broad-band inversion.

Velocity analysis and quality control from sub-stacks

We present a robust and accurate method of velocity analysis based on measuring any difference in timing between near and far offset stacks. Unlike semblance-based analysis this method identifies the correct moveout velocity of reflectors showing a class 2 AVO response with polarity reversal. If so desired, it can be run as a quality control in conjunction with semblance-based analysis and can flag the occurrence of a polarity reversal. Time shifts are measured by an optimally accurate maximum likelihood method. A criterion is established for detecting two special cases when the amplitude on either the near stack or the far stack is close to zero. A simple remedy is then to halve the substack fold. By including only the nearest or farthest offsets, the sub-stack sums only amplitudes dominated by the same polarity.