We can and should be using more prestack attributes to characterize unconventional reservoirs: An example from the Permian Basin

Stork, Christof (Land Seismic Noise Specialists) | Theisen, Sissy (Land Seismic Noise Specialists) | Vogler, Ray (Land Seismic Noise Specialists)


Although in conventional reservoirs a few seismic pre-stack attributes (e.g. AVO, Impedance Inversion) can directly correlate with the presence of oil and gas, this is not the case with unconventional reservoirs and their more complex rock properties. The rock properties that are important for unconventional reservoirs (e.g. brittleness, TOC, micro-porosity, clay content, stress state, fracture state) are very different than the important rock properties for a conventional reservoir (e.g. macro-porosity, fluid content). The relationship between the unconventional reservoir and the rock properties that influence seismic pre-stack attributes (e.g. bulk modulus, shear modulus, Poisson’s ratio, density, elastic stiffness parameters) are often indirect, complex, and vary from field to field. As a result, using pre-stack seismic attributes to characterize unconventional reservoirs is difficult and inconsistent.

We propose that resolving more (6-10 instead of 2-3) independent pre-stack seismic attributes are very helpful and may be a tipping point for using seismic data to characterize unconventional reservoirs. Because of the complex rock properties, the relationship of pre-stack seismic attributes to important reservoir properties may reflect some unknown complex physical processes or may even be an empirical relationship, which happens often in geology. One is much more likely to resolve this complex relationship with more pre-stack attributes because 4∧3 is 64 while 4∧10 is much greater, ∼1,000,000, where 4 represents the typical number of information bits we can resolve from a single pre-stack attribute and 3,10 represent the number of attributes. More attributes enable exponentially more precise reservoir characterization.

The key to accurately resolving 6-10 independent pre-stack attributes is recent advances that better address the significant noise in land seismic data. Using an example from the Permian Basin, we show that recent land acquisition and seismic processing advancements are available to achieve the data quality needed to resolve more pre-stack attributes. These recent advances that better address the land noise go by several names such as compressive sensing, simultaneous sources, massive fold, and machine learning/pattern-based noise removal, which are all offered by multiple contractors. We show the influence of these methods on a dense test survey in the Permian Basin.

This improved ability to deal with land noise is the key to reliably computing more pre-stack seismic attributes and resolving more complex rock properties in unconventional reservoirs.

Presentation Date: Monday, October 15, 2018

Start Time: 1:50:00 PM

Location: 209A (Anaheim Convention Center)

Presentation Type: Oral

  Country: North America (0.46)
  Geologic Time: Phanerozoic > Paleozoic > Permian (1.00)
  Industry: Energy > Oil & Gas > Upstream (1.00)