First Break - Volume 29, Issue 3, 2011

Huw James, Evgeny Ragoza and Tatyana Kostrova describe how the dramatically improved speed and memory of today’s graphics processing units are bringing a new dimension to the rendering of 3D seismic datasets.

Halina Jędrzejowska-Tyczkowska, Michał Malaga and Krystyna Żukowska present a method of quantitative prediction of velocity data from pressure data using geostatistical inversion.

Eivind Fromyr, John Brittan, Eivind Dhelie and Chris Page provide an update on how processing and imaging can contribute to the benefits of wide-azimuth seismic acquisition in geologically complex marine environments.

Friso Brouwer, Paul de Groot and Michael Kumpus show how a dense set of horizons enables seismic interpreters to extract more information from seismic data, using examples from the Middle East and the Canadian oil sands.

Liz Thompson and Camilla Oftebro review how geomodelling is maximizing the value of data in salt systems.

Prof Peter Hubral has more than once mused in the columns of First Break on the nature of creative thinking in science, based primarily on tenets of Chinese Tao philosophy which over many centuries have met with a surprising degree of corroboration from iconic scientists, philosophers, and writers. We recognize that this is not familiar territory for many of our readers, but nonetheless believe the ideas expressed deserve consideration by any scientist engaged in research and problem-solving activities.

Sequences of 4D seismic changes extracted over different time intervals from multiply repeated seismic surveys are correlated with the identical time sequences of cumulative fluid volumes produced from or injected into wells. Maps of these cross-correlations have previously been shown to produce a localized signal in the connected neighbourhood of individual wells. The technique is applied to the frequently repeated seismic surveys from the life of field seismic project in the Valhall Field, for which the 4D seismic signature is dominated by compaction-assisted pressure depletion. For these data, both acoustic impedance and time-shift attributes are found to have a remarkably consistent correlation with the well activity from selected groups of wells. Furthermore, maps of these results have sufficient fine scale detail to resolve interfering seismic responses generated by closely spaced wells and discrete zones of gas breakout along long horizontal producers. We conclude that uniting well data and 4D seismic data using our proposed method provides a new attribute for dynamic interpretation of the producing reservoir.

Sophisticated workflows have been designed to build reservoir models consistent with geology, production history, and seismic attributes. The seismic attributes are usually given as functions of depth on the vertical axis, which makes their integration relatively straightforward because reservoir grids depend on depth too. This approach requires that the velocity model used to convert the seismic attributes from the time domain to depth is fixed, whatever perturbation is subsequently applied to the geological model during the history matching process. Thus, the seismic attributes are converted only once from time to depth before performing history matching. As an alternative, we propose to design workflows that integrate seismic attributes given as functions of time instead of depth. This means that workflows must simulate seismic responses in the time domain. As fluid flow simulators yield saturations and pressures over the reservoir grid, the corresponding simulated seismic responses are first computed against depth. Then they are converted from depth to time using the high-resolution velocity model derived from the petro-elastic model. With this method, the depth to time conversion step is fully integrated in the history matching loop. A numerical example is presented to illustrate the potential of the proposed methodology.

Robust interpretation of seismic data relies upon identification and understanding of wavelet phase. Typically, well logs are used for the estimation of seismic wavelets, whereby the phase is obtained by forcing a well-derived synthetic seismogram to match a seismic trace. However, well logs are not always available, can predict different phase corrections at nearby locations and, due to sparse spatial sampling, cannot be used to accurately estimate phase variation in 3D. We introduce an extension to a statistical kurtosis-based phase estimation technique, proven able to estimate phase in agreement with seismic-to-well ties, without the use of well data. Our extension allows temporal and spatial phase variation to be estimated directly from the seismic data. Application of this method to a series of synthetic datasets demonstrates its ability to estimate true non-stationary phase, as well as apparent local phase anomalies, from clean or noisy data. Further tests on real seismic data show that the method can be used to locate apparent local phase anomalies caused by geology, such as multiple thin beds. We conclude that phase anomalies can be detected using statistical phase estimation techniques and could supplement the standard amplitude interpretation methods.