Geoscience Reference
In-Depth Information
fault segments. Detailed mapping of the relay ramps may be important in understanding
whether a fault will provide a seal to a prospective structure. Study of the way that
displacement of a horizon varies along a fault is needed to assess sealing capacity (see
chapter 4
), and is also a useful check on the correctness of the interpretation; along a
single fault, the displacement should increase smoothly from zero at the ends of the
fault trace to a maximum in the centre.
It is also possible to calculate the local curvature of the picked horizon (Roberts,
sign produced where the fault plane intersects the horizon on its up- and downthrown
sides. It may also be possible to relate curvature to fracture intensity, e.g. over salt swells.
The main problem in using curvature (which is a problem for dip and azimuth calculation
also) is to decide the length scale over which the attribute is calculated, distinguishing
larger-scale structural features from small-scale features that might be sedimentological
or might be noise (e.g. apparent reflector rugosity resulting from the autotracked pick
wandering up and down within a broad loop of low signal to noise ratio).
Another approach is to make an illumination display of the picked horizon. The
software calculates how the surface would look if seen from above when illuminated
by a light source from a particular direction. Usually the source ('sun') direction is
set to be near the horizontal ('low in the sky') so that subtle highs and lows in the
surface are picked out by the contrast between the bright surfaces facing the sun and
the dark shadows where surfaces face away from it. The effect is to emphasise those
topographic lineations that in map view trend at right angles to the sun direction.
Therefore, complete interpretation requires a number of illumination displays with the
sun in different directions. Better still is to have interactive control of the sun posi-
tion, with real-time updating of the screen display as the sun is moved around the
map. In this way the user can choose sun positions to emphasise particular features of
interest.
These methods depend on having an accurately picked horizon on a dense grid. If the
horizon of interest is not easy to autotrack, the interpreter will have to do a good deal of
editing and re-picking before he is able to use these tools. A different approach is to try to
recognise faults as discontinuities in the seismic trace cube, without necessarily having
any horizons picked at all. The basic idea is to calculate, over a limited time-gate, a
an implementation is the subject of an Amoco patent). The calculated similarity value
is posted in the seismic data cube at the centre of the trace window used to calculate it,
and the process is repeated for every trace in the seismic volume and for every possible
window start time; the result is therefore a complete data cube of similarity values.
Faults are revealed as planes of low similarity; they are best seen in a horizontal section
through the cube. The advantage over a simple time slice through the reflectivity data
is that the faults will be visible whatever their orientation; on reflectivity slices, faults
are often difficult to see where they run parallel to the strike of the bedding so that there
