Massive salt bodies in the deepwater Gulf
of Mexico can be both a blessing and a curse for oil companies exploring
the depths.
Yes, large salt-related structures can hold huge
accumulations of oil — but the same salt makes finding those big
fields a complicated and risky enterprise.
Consequently, the lure of big finds has prompted
companies to develop new techniques to reduce risk and improve seismic
imaging beneath and around salt bodies.
Pre-stack, depth migrated seismic data is a crucial
tool for today's deepwater explorers — and advancements in computing
technology and power has made it possible for seismic experts to
glean more and more information from data.
This is a story of how advanced depth imaging technology
helped one Gulf sub-salt operation.
"Sub-salt reservoirs are the ultimate goal of Gulf
of Mexico deepwater exploration," said David Kessler, president
of Houston-based Seismic City Corp., "and in the last few years
several new, large deepwater fields have been discovered using new
developments in imaging technology."
Kessler and his team recently worked with geologists
and geophysicists at another Houston firm, EEX Corp., to demonstrate
how advanced depth imaging technology can enhance the accurate and
successful definition of a new exploration prospect.
EEX acquired Mississippi Canyon blocks 975 and 976
and Atwater Valley block 8 in 1999 when Shell relinquished the acreage.
The firm believed the blocks had potential based on the Atwater
Valley 8 #1 well Shell Offshore drilled in 1997 to 20,700 feet.
Although the Cyclops well was deemed unsuccessful,
it encountered several reservoir quality sands in the Lower Pliocene
section.
"Based on 3-D data we had, we felt we could get as
much as 2,000 feet updip to those sands in Mississippi Canyon block
976 just to the north," said Marshall Thomsen, senior exploration
geophysicist with EEX.
The deepwater channel-sands and sheet-sands encountered
in the well were deposited in a lower-slope, fan depositional environment.
Eighty feet of gas pay was logged at about 14,400 feet, and other
sands had significant oil shows.
After acquiring the leases, EEX licensed 1998 vintage
pre-stack and post-stack time migrated seismic over the area from
Diamond Geophysical. Structural interpretation of this data was
EEX's first indication that the same sands at the Cyclops well could
be tapped updip at Mississippi Canyon block 976.
According to Thomsen, the prospect is a structural
trap below a salt/shale overhang. Massive salt walls form the trap
on the east and west flanks, while a rafted shale section — having
a remnant salt at its base — creates the trap to the north. Structural
dip is to the southwest.
Seismic data quality is quite good for mapping the
updip truncation of Pliocene-age sand found in the Cyclops well
against the rafted shale/salt weld, he added.
"The Larry prospect, as we call it, has potential
for multiple, stacked pays of Pliocene age with seismic amplitudes
increasing updip from the Cyclops well," Thomsen said. "Miocene
reflections seen on the seismic also subcrop updip against the rafted
shale section — (and) Miocene age sands produce at Shell's Europa
Field nine miles to the northeast."
Defining the Parameters
While the time domain seismic data EEX acquired revealed
a new prospect opportunity, the data was not imaged correctly in
areas associated with the thick salt on the east and west flanks
and the overhanging rafted salt/shale section.
Pre-stack depth migration modeling and imaging, applied
to better image the target structure, resulted in a more accurate
prospect definition and better well planning.
"One of the most important steps of the interpretation
process is to define the prospect area by extracting amplitude from
the seismic volume," Thomsen said. "In areas of complex geology,
the seismic image might contain both primary signal as well as seismic
noise. In areas where lithology is rapidly changing, 3-D pre-stack
depth migration is needed to correctly image target sands terminating
against a salt weld."
To help the technique to properly migrate the seismic
data in areas of strong lateral velocity variations, the team used
a modeling tool that predicts the wave patterns of the seismic energy
recorded during the acquisition phase, explained Jeff Codd, vice
president seismic technology for Seismic City.
A new modeling algorithm — called wave-front reconstruction
— simulated complex wave propagation from the surface to the sub-salt
target area. Due to the accuracy of the method, the resulting seismic
image had very little numerical artifacts.
Three-D pre-stack depth migration was applied in
such a way that the relative amplitude of the seismic data was preserved.
"This enabled us to successfully extract seismic
amplitude directly from the depth migrated volume," he said, "resulting
in a very clear definition of the updip limit of the prospect."
Stretching for an Answer
The reservoir quality sands at the Cyclops well ranged
in thickness from 50 to 200 feet, and a check-shot corrected synthetic
seismogram was generated from the sonic and density logs acquired
in the well.
This synthetic correlates with the seismic data and,
for mapping purposes, was used to identify reflections associated
with the thicker sands — many of which have an increase in amplitude,
updip from the Cyclops well, to their termination at the overhanging
salt/shale section.
At the Larry prospect, paleo markers identified in
the Cyclops well were input to the seismic time data and tied to
it using a velocity function derived from a check-shot corrected
synthetic. Stretching and squeezing was done to tie the synthetic
to the seismic using a log editing software.
A map generated from the seismic volume described
the prospective target, but the map had to be converted from time
to depth after imaging.
"Depth migration accuracy depends on the velocity
field input to the migration process," Kessler said. "This velocity
field is a detailed model of the subsurface geology that describes
the main geological units — in this case, a complex salt body embedded
in a sedimentary section."
Inputting an accurate velocity field to the imaging
process will produce a depth section that positions seismic events
close to their true depth.
"In cases of velocity anisotropy, the imaging algorithms
can use anisotropic parameters to more accurately model seismic
wave propagation in the subsurface," Kessler continued, "and therefore,
more accurately position seismic events in depth."
The advantage of direct mapping depth seismic volumes,
according to Kessler, is that no simplifications of the earth model
are done in order to tie the seismic data to known well data. The
resulting structure map constructed directly from the depth volume
is more accurate than the classic workflow of mapping in time domain
and then stretching the map to depth.
The synthetic log was converted to depth and correlated
to the pre-stack depth migrated data to determine any depth misties,
Kessler said.
This procedure also was used to tie in an Oryx well
drilled in Mississippi Canyon block 975, where depth misties ranged
from 20 feet at the water bottom to 200 feet at a depth of 20,400
feet.
Is It Salt or Shale?
In many cases of Gulf of Mexico seismic data interpretation,
scientists analyze high amplitude markers and determine if they
are related to salt or shale, according to Peter Harth, senior geophysicist,
and Glen Denyer, depth migration specialist, both with EEX.
At the Larry prospect it was imperative to differentiate
between a salt body with a distinctive base and a rafted shale body
on top of a salt weld.
"For the purpose of detailed well planning, it was
important to determine if the overhanging section is salt or shale
since the proposed well is designed to penetrate about 1,500 feet
of formation overhang," Denyer said.
A shale raft and any drilling problems associated
with it, including basal shear zones, must be isolated in a single
hole section, he added, and directional work in the raft and any
basal shear zone should be avoided due to potential instability
in pre-failed material.
"Determination of the raft, possible basal shear
zones underneath the raft and location of remnant salt section are
all partial constraints to determine where to set casing," he said.
The salt or shale question was resolved during the
depth imaging process: Two main operations completed during the
model-building phase — the construction of the salt body, and the
construction of the velocity field around the salt — produced "interval
velocities" directly related to lithology, according to the EEX
scientists.
This technique — a new velocity analysis based on
a depth migration algorithm — gives processors the ability to define
velocity variations within geologic formations with a 1 percent
margin of error.
At the Larry prospect shale units were differentiated
from the salt body by careful analysis of 3-D pre-stack depth migration
image gathers. Special effort was made to carefully analyze the
velocity above the high amplitude marker, resulting in a very slow
velocity field in this area.
"The resulting interpretation of the velocity field,
together with the seismic image, concluded that the geology is of
a rafted shale section located on top of a salt evacuation weld,"
Denyer said, "with a remnant salt of about 1,200 feet thick at its
base."
Mission Accomplished
Four different seismic volumes were used during the
interpretation of the Larry prospect: The original 3-D post-stack
time migrated data; a 3-D post-stack depth migrated volume; a 3-D
pre-stack time migrated volume; the 3-D pre-stack depth migrated
data.
The pre-stack depth migrated volume definitely provided
superior formation correlation, discontinuity resolution and deep
amplitude preservation when compared to the previous processing,
according to both Thomsen and Michael Padgett, EEX's vice president
of Gulf of Mexico exploration.
"Again, the issues for the Larry prospect are target
placement and hazard avoidance," Thomsen said. "With the 3-D pre-stack
depth migrated data, the location and thickness of the rafted shale
section are well constrained, which allows for casing-setting above
and below this interval."
In drilling down from and along the basal salt, he
continued, the pre-stack depth migrated image allows well placement
to be significantly updip of the Cyclops well, while minimizing
the probability of crossing into salt.
"The vertical error for target placement is expected
to be plus or minus 200 feet," Thomsen said. "Also, as a well is
drilled, the pre-stack depth migrated volume can be continuously
tied and depth-updated, which decreases the uncertainty during the
drilling process."
As Padgett pointed out, the two primary goals of
any seismic project are to unravel the geology and help reduce the
drilling risks.
"Today nobody is going to drill around these shale
or salt masses without 3-D pre-stack depth migrated seismic," Padgett
said, "especially in the deepwater Gulf of Mexico, where a drilling
disaster on a planned $20 million well ends up costing $50 to $80
million. Wells of that magnitude demand the best possible imaging
technology."
Three-D pre-stack depth migration technology was
originally developed to image seismic data in areas where time domain
imaging failed, according to Denyer and Harth. Today's depth processing
technology can produce seismic depth volumes that bring many advantages
to deepwater exploration.
"We foresee that further development in depth imaging
and modeling technology will improve our ability to interpret seismic
images and understand seismic amplitude," Denyer said, "and will
help in lowering the risk of deepwater exploration."