Resource plays are one of the most active exploration arenas in the world. Many of them consist of fractured formations containing large amounts of gas. The challenge with many of these prospects is not just discovery but also production and exploitation.

Although certain attributes have proven successful at predicting the existence of natural fractures in the formations, they have fallen short of indicating the orientation of such fractures. Fracture orientation is an important piece of information required for successful steering of a well path, especially when dealing with highly deviated or horizontal drilling. With the advent of wide azimuth acquisition surveys, we find ourselves in a position to do the proper data processing in the azimuthal domain.

New imaging and analysis tools allow for the extraction of fracture information in the reservoir. The technology is based on the use of offset vector tiles (OVT), azimuthally preserving migration and migrated offset gathers.

This geometry and Rose diagram correspond to the table above. (Images courtesy of Geotrace)

A well-sampled wide-azimuth dataset

The process starts by analyzing the acquisition geometry and determining if it contains the offset and azimuthal fold required for the formation of OVT. In addition to the standard geometric considerations regarding bin size and time sampling, the relevant domains that need to be well-sampled are offset and azimuth. For a hypothetical survey with the dimensions in the table below, we can see (Figure 1) that the azimuthal/offset coverage is appropriate for all azimuths and offsets up to 15,000 ft (4,575 m, black circle).

Offset vector tiles, migrations, and gathers

Hybrid gathers are single-fold 3-D cubes of data formed by a resorting of shot and receiver lines into cross-spreads. The sizes and shape of these objects is defined by the size of the recording patch. Processing hybrids gathers is well understood and offers many advantages, primarily for 3-D prestack noise elimination approaches, and it has been documented extensively (Stein & Langston 2007).

OVTs are a natural extension of hybrid gathers (Vermeer 2002, 2003, & 2007), and they provide a set of constant offset-constant azimuth datasets that can be migrated using a special-purpose Kirchhoff algorithm that preserves the azimuthal information. The resulting common image gathers are called offset vector gathers (OVG).

Following the standard methodology for measuring anisotropic parameters from migration, the process starts by performing an isotropic but azimuth preserving migration (Figure 2). The OVG’s are then sorted into azimuth and offset gathers for analysis. The deviation from flatness observed in the gathers is indicative of the existence of anisotropy. Here lies the beauty and power of this process.

The measurable anisotropy is due to the combined effect of the layering in the earth, which produces vertical anisotropy, known as vertical transverse isotropy. The vertical fracturing of the reservoir produces horizontal-type anisotropy, know as horizontal transverse isotropy (HTI). If both exist simultaneously, we call it orthorhombic anisotropy (Jenner et al 2001).

This is the OVT Workflow needed for the construction of FracMaps and schematics of the OVT is migration.

FracMap volume construction

By properly accounting for anisotropy of all kinds, the migrated gathers can be made flat (in the azimuth and offset domains) and a final interpretable section can be produced. The methodology used to achieve this is not the focus of this paper and will be dealt in a separate publication. Our focus will be exclusively on how to extract these anisotropic parameters and exploit the richness of information contained within them, and how to translate it into fracture information.

The moveouts observed in the azimuthal and offset domain (Figure 3) can be measured and used to infer the desired fracture intensity and orientation. This is done by essentially unfolding the OVG into a set of gathers properly sorted in offset and then azimuth.

OVGs sorted into offset and azimuth display HTI anisotropy for every azimuth.

Unlike normal moveout, which depends exclusively on offset and vertical root mean square velocity, orthorhombic anisotropy is a function of the vertical velocity (VV) and two horizontal velocities normally referred to as (VHfast, VHslow). These three describe an ellipsoid in velocity space. It is the size and orientation of this ellipsoid that determines the fracture intensity and orientation (Figure 4)

For simplicity of argument we will assume that there is no vertical anisotropy present so that the ellipsoid becomes an ellipse and we only need to deal with one “moveout.” Figure 5 shows a portion of an OVG gather for a particular offset. The curved nature of the moveout is readily observed and can be measured. This is done by fitting the velocity ellipse to the measured delays (?t) and associating the “orientation” of the ellipse (?) with the direction of the fast propagation and consequently the orientation of the fractures

This velocity ellipsoid describes orthorombic anisotropy.

By performing this ellipse (or ellipsoid) fitting for every OVG in the survey, a volume of directional vectors with varying magnitudes and directions is generated. We call it FracMap. Figure 6 shows a portion of a time slice extracted from one of the FracMaps generated. It shows a set of faults almost describing a V shape. After closer inspection we can see that the left arm presents a set of fractures that are perpendicular to the fault direction, while the right arm shows fractures aligned with the fault direction. These considerations can dramatically affect the drilling decisions.

Fracture intensity and orientation are computed.

OVT are the most efficient way to do basic processing, including noise attenuation, interpolation, regularization, imaging, velocity determination, amplitude versus offset, and rock property inversion while preserving the azimuthal information.