D

isplay of seismic and well

data together can strengthen
the interpretation of both.
The complex field of seismic data
acquisition requires reliable software
that can link disparate data and
disciplines together, while offering
computer assisted tools to help
users acquire crucial geological cross
sections. Marker correlation for
horizontal and highly deviated wells
can be supported by using vertical
traverses through well markers or
by correlating well markers in 3D
views, both methods allow the well,
seismic and interpretation data to
be spatially correlated and avoid the
problem of losing positional reference
when converting deviated wells from
measured depth to true vertical depth
or onto vertical sections.
Exploration and asset development
teams are composed of multiple
disciplines which use different tools
to analyse different data types in
different disciplines. This means seismic
data has not been readily available to
petrophysicists and that geologic cross
sections have been unavailable to seismic
interpreters.
But 3D visualisation systems have
been successful in capturing 3D seismic
interpretation tasks.
As more data types such as well,
satellite imagery and downhole
measurements have been added to
these displays, for the benefit of seismic
interpreters, geologists and petrophysicists
have also benefitted.
Seismic data and seismic analysis tools
are now readily available from 3D and
2D seismic surveys for use by geologists
building cross sections, petrophysicists
analysing log curves and drillers planning
wells.
Modern systems can add data from
multiple 2D and 3D seismic surveys; data
from multiple well data bases and multiple
seismic interpretations to a single section
or 3D view.
This makes comparison of different
interpretations far easier than before, and
enables the construction of consolidated
interpretations across many interpreters as
well as across many disciplines.
We can now see seismic interpreters
using the cut-offs and cross plots of
petrophysicists; geologists using the
picking and drawing tools of geophysicist,
and all disciplines performing both
structural and stratigraphic interpretation,
using facies analyses of well and
seismic data and using flattened and
proportionately sliced displays.
The use of common section and 3D
views enables all disciplines to benefit
from the same developments and share in
the costs, but more importantly, it enables
different disciplines to collaborate in the
same scenes and lowers the barriers
between them caused by the use of
different tools.
This enables much better
communication between various
disciplines in exploration and asset
teams, which may result in a much better
synthesis of all available data.
This unified interpretation means
anomalies become easier to see,
resolve and integrate, or throw doubt
on the interpretation. Many plausible
interpretations in any one discipline can
be vetoed once all the available data is
integrated.
Next generation software
The seismic components of next
generation software can be characterised
by the increasing availability of computer
assisted tools. The most obvious of these
has been 3D seismic horizon trackers.
These tools have been augmented by 2D
seismic horizon pickers, computer assisted
fault pickers, and various real time
tessellation schemes that bring modeling
into the interpretation workflow.
These have been joined by log
correlation methods that can follow
geologic rules, filtering techniques that
can enhance continuity or discontinuities
special feature: Seismic
Correlating well markers
with the aid of seismic
interpretation
Huw James and Taoufik Ait Ettajer of Paradigm, exploration and production geoscience software manufacturer,
illustrate the possibility of correlating well markers with the aid of seismic interpretation and seismic data in
section and 3D views.
www.pipelinedubai.com
44 Pipeline 05/2010
for horizon and fault picking respectively;
and of course the myriad of attributes,
facies analysis and inversion schemes that
can process multiple seismic channels
and data sets into various physical
measurements such as Fluid Factor or
enhance particular aspects of the seismic
data.
Geologists and petrophysicists are now
benefitting from similar computer assisted
methods that allow thousands of faults
to be extracted from 3D data sets and
then analysed for orientation between
successive unconformities.
The statistics resulting from massive
amounts of data can reveal quite subtle
trends in the orientation of faults over
time. With partial automation, large
amounts of well log data can be analysed
similarly and subtle relationships of curve
data are found for particular geologic
intervals and areas. This can lead to much
better and precise estimates of rock
properties. These tools are making today
an exciting time for earth scientists.
Down to Earth
To get practical, the simplest way of
showing the
benefits is to
define lines
of section or
traverses between
wells on a base
map.
Once this
traverse is defined,
3D seismic data
can be extracted
along the traverse
to indicate changes
in structure and
stratigraphy
between the wells.
This was one of the first immediate uses
of 3D seismic interpretation systems and
it greatly simplified fault interpretation
from well data. The main difference today
is that the seismic data is live for the
geologists to use. It can be filtered with
structurally oriented filters that preserve
discontinuities or it may be pre-stack
gathers along the traverse which can be
examined for AVO effects. Dead seismic
has been replaced by live seismic that
can provide better
integration with
well data.
The ease of
generating all
available data along
any well traverse
means interpreters
can interpret in
many different
azimuths through
the survey, and
that they are no
longer constrained
to the acquisition
directions of in-
line and cross
line. Traverses can
be dip and strike
or radial and tangential around circular
features such as salt domes.
This case is quite simple, but obviously
the different well tracks can be used to
display biostratigraphic, lithostratigraphic,
chronostratigraphic, computed log,
perforation, pressure test, fluid data and
seismic synthetics.
If the wells are not vertical, the
interpreter can define the well traverse
through any marker including the bottom
hole marker in order to keep the vertical
curtain assumption of a simple cross
section as close to reality as possible.
So long as the wells are not horizontal
there will be a limited vertical plane that
is close to the well bore for any particular
depth or time interval. Such cross section
displays enable well data to be visually
correlated with the seismic data which
in turn drives the correlation of well data
from well to well.
There are many instances in Figure
2 where a particular log response
correlates to large seismic amplitudes with
significant structure from well to well. For
example, a yellow peak in the porosity
log on well KG-5X at @ 1650 ms can be
relatively reliably correlated with a similar
peak in well KG-2AX @ 1300ms.
Without the seismic data this correlation
would be difficult to carry across well KG-
1X with its different shape of the gamma
ray curve.
Conversely, the injection of the well data
special feature: Seismic
Figure 1 shows a simple time structure base map
with fault outlines and a simple traverse through
five wells.
Figure 2 shows a sample cross section with well log panels inserted that
combine seismic data and seismic interpretation with the well data and a
stratigraphic column.
www.pipelinedubai.com
45 Pipeline 05/2010
into the seismic traverse also illuminates
the seismic interpretation and causes
some re-appraisal of the seismic data
around well KG-2AX.
You can easily see how seismic
interpretation methodology could switch
from one of interpreting and propagating
horizons from views in in-line and cross
line space, skipping use of arbitrary
traverses for propagation and settling
on views of well traverses so that
interpretation data will be more closely
tied to well data.
3D structure and views
Section views are sufficient to interpret
layer cake geology with simple dip and
strike directions and vertical wells. If the
geologic targets are 3D in nature, for
example channel fan complexes, reefs or
complex structures, then 3D views can be
very beneficial for the interpreter.
3D visualization enables the interpreter
to effortlessly see the structure, and then
use their brains to understand the details
in the data, rather than to be fully engaged
in constructing a mental image of the
structure.
3D views also allow groups of
interpreters to collaborate more easily.
When each interpreter makes individual
images of a 3D structure, it is difficult to
ensure everyones
images are aligned.
But, when the
structural image
is generated
by computer,
alignment is
forced upon the
interpreters.
The computer
generated image
has to be visually
plausible to satisfy
most geologists
- because their
brains are well
practiced at
creating their own
3D images and
implausible or
inaccurate displays can be painful for
such an audience!
The interpreter automatically
understands when part of the traverse
is along strike and when part of the
traverse is perpendicular to strike. This
understanding would be much less direct
for an interpreter viewing a section. The
azimuth of the section with respect to
strike would be unknown. The structural
reference is extended out into the view
with the aid of a partially transparent
horizon surface. Here the 3D view of the
data aids the interpreter because of the
3D geometry of the
structure.
Channel-levee-
fan complexes
are also much
better viewed in
3D where the full
3D structure can
be seen in a single
view rather than
in multiple section
views where the
3D structure has
to be built up in
the mind of an
interpreter. If the
section views are at
unfortunate angles to the structure it may
escape recognition when viewed slice by
slice.
Deviated and horizontal wells
Horizontal wells are difficult to correlate
from 1D log displays. Curves from two
wells that are close in the subsurface
may be far apart in measured depth
because of differences in the two borehole
trajectories.
It is not possible to convert horizontal
wells to true vertical depth without losing
large quantities of data in the horizontal
sections. It is also difficult for the brain
to construct internal 3D images of
boreholes relative to the 3D structure if
the boreholes are not parallel or otherwise
simply arranged. 3D views allow the
boreholes to be placed correctly in 3D
space and the 3D structure can be added
to the view with 3D views of horizon and
fault surfaces.
Adding extracted seismic data to the
surface joining two well bores enables
the intervening space to be inspected for
faults or changes of stratigraphy indicated
by the change in the pattern of seismic
data.
Vertical curtains of seismic traverses can
also be added to give the interpreter extra
clues about the structure and stratigraphy
of the volume. A sandy bed indicated by
a blue shade of the gamma ray log can
be correlated between the wells and
Figure 3 shows a 3D view across a major fault with about 600 metres of relief.
Figure 4 shows two almost horizontal wells with gamma ray log curves and a
seismic well section extracted from a 3D data volume.
special feature: Seismic
www.pipelinedubai.com
46 Pipeline 05/2010
the intervening space can be inspected
for indications of seismic or sub-seismic
faults. In this case there is a hint of a
subtle fault.
Summary
If you view the figures - and try to
observe your brain in action - you should
be able to see your mind interpreting the
well data, and then inspecting the seismic
data; followed by the reverse process of
interpreting the seismic data and then
trying to understand its relationship to the
well data.
This deepens the seismic interpretation
from one of picking structure to one of
seeking reservoir material, in this case
fairly thin sands. Then, when the seismic
and well data do not correlate well, the
brain tries to remove the conflict.
On a workstation the interpreter would
take a look at an alternative view to seek
resolution of any problem. This is not
possible from hard copy.
So just placing the data in spatial
proximity helps correlation of both
seismic to seismic, well to well and well
to seismic. Placing the data in a 3D view
is of great help when there is significant
structure and is hugely beneficial if the
wells are horizontal.
The data used in this paper is from
offshore Indonesia courtesy of Clyde
Petroleum.
special feature: Seismic
www.pipelinedubai.com
47 Pipeline 05/2010