Page 4 - i1052-5173-27-2

Impact of seismic image quality on fault interpretation uncertainty

Juan Alcalde, Geology and Petroleum Geology, University of Aberdeen, School of Geosciences, Kings College, Aberdeen, AB24 3UE,
UK, and School of GeoSciences, University of Edinburgh, West Mains Road, Edinburgh, EH9 3FE, UK, juan.alcalde@abdn.ac.uk;
Clare E. Bond, Geology and Petroleum Geology, University of Aberdeen, School of Geosciences, Kings College, Aberdeen, AB24
3UE, UK, clare.bond@abdn.ac.uk; Gareth Johnson, School of GeoSciences, University of Edinburgh, West Mains Road, Edinburgh,
EH9 3FE, UK, g.johnson@ed.ac.uk; Jennifer F. Ellis*, Midland Valley Exploration Ltd, 2 West Regent Street, Glasgow, G2 1RW, UK,
ellisj11@cardiff.ac.uk; and Robert W.H. Butler, Geology and Petroleum Geology, University of Aberdeen, School of Geosciences,
Kings College, Aberdeen, AB24 3UE, UK, rob.butler@abdn.ac.uk

ABSTRACT faults. The resulting interpretations of fault Bahorich and Farmer, 1995). However,
patterns are used to infer a wide variety of there are many other explanations for
Uncertainty in the geological interpreta- tectonic properties—for example: estima- reflector termination, some geophysical
tion of a seismic image is affected by tions of upper crustal stretching during (e.g., noise, processing effects, anomalous
image quality. Using quantitative image lithospheric extension (e.g., Kusznir and changes in velocity) and some geological
analysis techniques, we have mapped dif- Karner 2007); kinematic connectivity and (e.g., depositional facies changes, channels,
ferences in image contrast and reflection stretching directions (e.g., Baudon and unconformities), that are not always easy
continuity for two different representations Cartwright, 2008); and polyphase reactiva- to distinguish, so there are ambiguities in
of the same grayscale seismic image, one tion and inversion (e.g., Underhill and fault interpretation. Subtle differences in
in two-way-time (TWT) and one in depth. Paterson, 1998; Badley and Backshall, fault interpretation introduce changes in
The contrast and reflection continuity of 1989). Fault interpretations are important the geometric characteristics of the faults
the depth image is lower than that of the components in the prediction of hydro­ (e.g., throw, heave), with, for example,
TWT image. We compare the results of carbon reservoir volumes in structural traps, impact on the determination of stretching
196 interpretations of a single fault with and in forecasting the integrity and perfor- factors for sedimentary basins. For basins
the quality of the seismic image. Low con- mance for structurally complex reservoirs in a late stage of being explored, 3D seis-
trast and continuity areas correspond to a (e.g., Richards et al., 2015; Yielding, 2015; mic data are often employed because they
greater range of interpreted fault geom- Wood et al., 2015; Freeman et al., 2015). generally provide a higher spatial resolu-
etries, resulting in a broader spread of However, in publications, faults are com- tion and geometric continuity compared
fault interpretations in the depth image. monly shown as single, deterministic with even closely spaced grids of 2D seis-
Subtle differences in interpreted fault interpretations—even though there are mic profiles (Cartwright and Huuse, 2005;
geometries introduce changes in fault uncertainties in these seismic interpreta- Gao, 2009), but there can still be signifi-
characteristics (e.g., throw, heave) that are tions that will impact the application of the cant uncertainty in structural interpreta-
critical for understanding crustal and litho- interpretation. A single seismic image can tion. Regardless of the development of 3D
spheric processes. Seismic image quality comprise a range of interpretations with seismic methods, 2D data continue to
impacts interpretation certainty, as evi- intrinsic probabilities (Bond et al., 2007; underpin regional tectonic studies and
denced by the increased range in fault Hardy, 2015). Despite the importance of frontier basin exploration (e.g., Platt and
interpretations. Quantitative assessments fault interpretations, remarkably few publi- Philip, 1995; Thomson and Underhill,
of image quality could inform: (1) whether cations or indeed training materials explain 1999; Gabrielsen et al., 2013). Much of the
model-based interpretation (e.g., fault how faults are interpreted on seismic understanding of fault geometry is based
geometry prediction at depth) is more robust reflection profiles or discuss the uncertain- on heritage 2D data from the 1980s (e.g.,
than a subjective interpretation; and ties in the interpretations. Here we explore Freeman et al., 1990), even if enhanced by
(2) uncertainty assessments of fault inter- how image quality impacts fault interpre- subsequent 3D studies (e.g., Cartwright
pretations used to predict tectonic processes tation, using outputs from an interpretation and Huuse, 2005). Furthermore, training
such as crustal extension. exercise. materials in fault interpretation (e.g.,
Shaw et al., 2005), as well as knowledge-
INTRODUCTION Faults may be characterized as quasi- sharing resources (i.e., books and arti-
planar features that offset geological mark- cles), are chiefly two-dimensional, pre-
Interpreting seismic reflection data is ers. It is rare that the fault surfaces them- sented in paper or on computer screen. In
the principal approach for obtaining a selves generate seismic reflections. summary, 2D interpretation is a funda-
detailed understanding of the geological Therefore, on seismic images, fault geom- mental and important part of most seismic
structure of the subsurface. Central to etries are established chiefly by linking the interpretations irrespective of whether the
these endeavors is the ability to trace terminations of stratal reflectors (e.g.,

GSA Today, v. 27, no. 2, doi: 10.1130/GSATG282A.1
*Current address: Cardiff University School of Earth and Ocean Sciences, Main Building, Park Place, Cardiff, CF10 3AT, UK.

4 GSA Today | February 2017