The European Association of Geoscientists and Engineers is one of the leading geoscientific organizations in the petroleum industry today. Each year thousands attend its annual conference and exhibition. At these events the majority of papers and exhibits deal with seismic acquisition, processing, and interpretation — indicating that seismic technology is experiencing a golden age.
Advances in seismic technology have indeed been phenomenal, especially in the domains of seismic acquisition and processing. Denser sampling in space and time, better algorithms to remove unwanted noise, and ever-improving imaging allow geoscientists today to construct increasingly accurate geological models on which important exploration and production decisions are based. Progress has also been made in the domain of seismic interpretation through the introduction of sophisticated attributes and innovative visualization techniques. However, it is my belief that in the field of seismic interpretation we are yet to enter the golden age. In fact the best is yet to come. Let’s look at what has been achieved in seismic interpretation to date, and the range of future possibilities.
To unlock the hidden treasures in seismic data today we need to introduce the concept of geological age into the interpretation process. Time in geology is a poorly understood feature. Professor Salomon (Salle) Kroonenberg, an éminence grise of Dutch geology, wrote a very interesting book about the subject, The Human Scale: The earth ten thousand years from now. Salle says time in geology manifests itself in three ways: as a flow, as a pulse, and as a wave. For example radioactive decay, the expansion of the universe, and evolution are processes in which time manifests itself as a flow — what Stephen Jay Gould called time’s arrow. Earthquakes, meteor showers, eruptions, floods, and extinctions are examples of time as a pulse. The third manifestation, time as a wave, is about cyclical processes. Examples are continental plates breaking apart, colliding, and breaking again; ice ages alternating with warmer periods; and sedimentation processes controlled by Milanković parameters.
In seismic interpretation we are primarily concerned with time as a wave. In the 1970s Pete Vail and his colleagues at Exxon started to map seismic reflection patterns. They observed cyclical patterns that could be explained in terms of sedimentation processes. Moreover, they discovered that seismic reflectors are the first order approximations of geological timelines. In other words, mapping horizons that follow seismic reflectors is essentially equivalent to mapping geological time. A new interpretation technique called seismic sequence stratigraphy was thus born. The technique helped improve our understanding of depositional systems and has been used ever since to find stratigraphic traps. Due to the lack of supporting software algorithms, seismic sequence stratigraphy, however, never reached its full potential with interpretations carried out manually in what was a cumbersome and time-consuming process.
Recently, however, a new group of semi-automated seismic interpretation techniques have emerged that aim to generate fully interpreted seismic volumes. The algorithms behind these techniques all share in common the fact that they try to correlate seismic positions along geological timelines.
- Tracy Stark’s ‘age’ volume assigns a value representing relative geological time to each seismic sample position. The age assignment is based on correlating instantaneous phase signals from trace to trace.
- The PaleoScan software from French startup company Eliis builds a geological model roughly on the scale of the seismic sampling by connecting each seismic event to the most probable neighbouring event.
- Volumetric flattening, Chevron’s proprietary technology by Jesse Lomask, uses similarity-correlated surfaces to flatten the seismic volume or attributes of it. The flattened volumes are known as Wheeler cubes.
- dGB’s HorizonCube algorithm correlates timelines in the pre-calculated seismic dip field. The tracked surfaces are stored as a dense set of mapped horizons called HorizonCube.
Fully interpreted volumes are used, among other things, to assist in well correlations, in unravelling depositional histories, and finding stratigraphic traps using sequence stratigraphic interpretation principles. Other applications include detailed geological model building and improved seismic inversion and reservoir property prediction schemes that start from more accurate low frequency models. In addition, geohazard interpretation, geo-steering, and the finding of sweet spots in unconventional plays have all benefitted significantly from fully interpreted volumes that can add value to the seismic data.
While enormous progress has been made, it’s clear that the current group of global interpretation techniques is still evolving. Interpretations covering different geological settings such as passive margins and carbonate platforms have already been published. But as the technology matures, additional applications and more complex settings will be put to the test. What is clear is that we are on the verge of entering a golden age of seismic interpretation. Watch this space!