Vous êtes sur la page 1sur 7

THE ESTIMATION OF FORMATION PORE PRESSURE FROM ACOUSTIC

VELOCITY AND TOMOGRAPHIC MIGRATION VELOCITY ANALYSIS FOR


AN AREA WITHIN WEST CAMERON, OFFSHORE LOUISIANA
John Smythe, Dennis Loren, Paul Docherty, Rick Box
OVERVIEW
Pore-pressure prediction has applications in
analysis of hydrocarbon migration and seal
capacity, determination of reservoir drive and
connectivity, analysis of drilling hazards (i.e.
shallow water flows) and well casing design.
Given the water depths, drill depths and well
costs that challenge operators in the Gulf of
Mexico and around the world, pore-pressure
prediction should be an integral part of prospect
evaluation and well planning.

Using this relationship, seismic velocities can be


used to predict pore pressure.
Clearly, successful seismic pore-pressure
prediction depends on reliable well log data,
careful pressure-velocity calibration, and dense
and accurate seismic velocity analysis. In this
study, a quality pore-pressure prediction was
assured through five key steps:
1.
2.

In four blocks in the West Cameron area of the


Gulf (Figure 1), Fairfield Industries produced a
3-D prestack depth migration using velocities
derived by a technique known as tomographic
migration velocity analysis (MVA). In the same
area, Loren and Associates undertook a nine well
study to relate log-derived acoustic velocity to
formation pore-pressure. The relationship, when
applied to Fairfield Industries tomographic
velocities, can predict pore pressure throughout
the 3-D depth migrated seismic volume.
INTRODUCTION
The primary variable that relates velocity to fluid
pressure is effective stress. Effective stress is the
amount of overburden pressure that is supported
by the rock matrix. The effective stress vs.
compressional transit time trend depends on
lithology and must be locally calibrated.
Overburden pressure is the combined weight of
the water column above the sea floor, the rock
matrix, and the fluid in the pore space above the
zone of interest. Overburden pressure is a logderived property obtained from integrating the
carefully edited and continuous density log for
the entire length of the well bore.
Pore pressure is equal to the overburden pressure
minus the effective stress. Effective stress has
been found to be dependent upon compressional
acoustic velocity after accounting for variations
in lithology and depth (pressure-velocity
calibration). Correctly edited sonic and density
logs define the specific fluid pressurevelocity
relationships applicable to the project area.

3.
4.
5.

Multi-well editing (Log Purification) and


calibration.
Derivation of overburden pressure and
effective stress in calibration wells.
Determination of velocity/effective stress
transform.
Tomographic migration velocity analysis.
Application of velocity/effective stress
transform to seismic velocities.

LOG PURIFICATION
For many wells sonic and density logs are often
unavailable or unreliable due to borehole
conditions (rugosity, stress relaxation, alteration
during prolonged periods of exposure to drilling
fluids, etc.). Accurate and continuous sonic and
density logs are fundamental to pore-pressure
prediction.
In order to overcome logging
problems, Loren & Associates developed a
multi-well, multi-log editing process called Log
Purification. The process uses redundancy
between log curves to produce sonic, density,
and pressure curves representative of undisturbed
conditions.
The
Log
Purification
technique
uses
mathematical regression over all zones of valid
well log data in all wells to deduce,
simultaneously, relationships between functions
of compressional velocity, density, fluid
pressure, lithology, deep resistivity, and depth.
A simulated version of any log curve (sonic,
density, etc.) can then be constructed from the
other curves. Intervals of questionable log data
can be recognized and replaced with calculated
data, while intervals of missing data may be
filled in. This methodology is an improvement
on traditional log editing techniques that fail to
recognize calibration differences between wells.

Figure 2 is an example of a log before and after


the purification process. Figure 3 is a typical log
curve set generated for each calibration well.

depth can be converted into errors in traveltime


through the earth model. The model is then
updated to minimize these traveltime errors.

PRESSURE-VELOCITY CALIBRATION

Tomography is the approach we use to update


the model. Given the traveltime errors between
many thousands of imaged points at depth and,
typically, equally many seismic sources and
receivers on the earths surface, a large system of
equations is solved simultaneously.
The
equations describe the paths that sound takes
through the earth, while the unknowns are the
velocities of the rocks through which the sound
has passed. It is not uncommon for this system
to contain tens, or even hundreds, of millions of
equations. We obtain a so-called global solution,
that is, the entire earth model is updated, unlike
the more common layer-stripping or top-down
approaches.

Formation fluid pressures are not accurately


known at most points in the well. Fluid
pressures within shales and untested sands must
be inferred from drilling performance, log
relationships, and fluid pressures measured
within adjacent permeable formations. One
product of the Log Purification process is a
continuous fluid pressure gradient profile.
Compressional velocities in a given study area
depend on effective stress, lithology, and depth.
Traditional
approaches
to
pore-pressure
prediction (Eatons method, for example) depend
on an extrapolation of velocity trends in
normally pressured rocks to approximate depth
effects. Constant lithology, usually shale, is
assumed in order to arrive at velocity as an
approximate function of effective stress. The
extrapolation can be problematic if the data is
poor or the distance too great. In addition, local
geologic conditions, such as unconformities,
may interrupt petrophysical continuity.
Since this method differentiates based on
lithologies an interpreter may study the effects of
various lithology assumptions on the porepressure prediction. End member cases of pure
shale and pure sand as well as an estimated case
based on lithologies of nearby wells offer the
advantage of visualizing and dealing with
uncertainties in pressure prediction caused by
uncertainties in lithology.

TOMOGRAPHIC MIGRATION
VELOCITY ANALYSIS
Conventional seismic velocity analyses assume
that velocity varies slowly both laterally and in
depth; or worse still, that the earth is composed
entirely of flat-lying layers. Resolution obtained
from these methods is often too low for accurate
pore-pressure prediction. Tomographic MVA, on
the other hand, correctly accounts for distortions
in seismic waves brought about by rapid
variations in geology. The basis of the technique
is straightforward enough: A point in the
subsurface of the earth should image at the same
depth when seismic waves from different offsets
are depth migrated. Deviations in the imaged

EXAMPLE: WEST CAMERON,


OFFSHORE LOUISIANA
We selected a four block area in West Cameron
from Fairfield Industries speculative data library
for prestack depth migration. The migration
velocities were extracted from the data volume at
the calibration well locations. Figure 4 is a
comparison between the well velocity and the
extracted MVA velocity after several iterations
of tomography. The MVA produced a smoother
velocity but matched the overall trend of the well
velocity. It is likely that further iterations of
tomography, targeted at data from the
neighborhood of the pressure transition at 11,000
ft., would improve the match.
Figures 5 and 6 are inline extractions from the
velocity model and resulting pore-pressure
volume through one of the calibration wells. A
sonic log and fluid pressure gradient curve
clearly show a pressure transition at 11,000 ft.
This abrupt pressure transition occurred when
the well bore intersected a pressure sealing fault.
The line also shows a pressure anomaly in the
productive fault block.
The high pressure
anomalies associated with the field pays can
enhance
understanding
of
hydrocarbon
distribution throughout the reservoir.
CONCLUSIONS
The problem of relating formation fluid pressure
to log derived velocity, and by extension to
seismic velocities, requires consideration of
several important points:

1.

2.

3.

Formation fluid pressures are not


accurately known at most points in the
well.
Wireline formation tests, drill
stem tests and production tests are
absolute measurements of fluid
pressures but only available in discrete
and isolated points in the well.
However, fluid pressures within shales
and untested sands can be inferred by
log relationships during the Log
Purification phase.
Recorded density and acoustic transit
time logs are often not available
throughout
the
entire
borehole.
Additionally, recorded data may be
inaccurate due to borehole-related
problems
or
calibration
errors.
Defining pressure/velocity relationships
and overburden pressure profiles from
acoustic transit time and density logs
requires careful log reconstruction.
The relationship between compressional
velocity and effective stress depends on
depth and lithology.
Traditional
methods assume constant lithology,
making the task of interpreting them
problematic. Traditional methods also
extrapolate velocity trends based on the
normally-pressured portion of the well,
which can lead to error.

4.

The quality of the pore-pressure


prediction is directly relatable to the
quality of the seismic velocity analysis.
The tomographic velocity method used
by Fairfield is superior to traditional
migration velocity analysis.

5.

Using Fairfield Industries tomographic


MVA and log purification and
calibration techniques developed by
Loren and Associates we successfully
produced a pore-pressure volume in an
area in West Cameron, offshore
Louisiana. The volume can help predict
drilling hazards, pressure distribution in
a reservoir, and fault seal capacity.

More Information:
John Smythe
Fairfield Industries Inc.
14100 Southwest Freeway, Suite 100
Sugar Land, TX 77478
Phone (281) 275-7737
Dennis Loren
Loren and Associates
4600 Hwy. 6 North, Suite 200
Houston, TX 77084
Phone (281) 861-6503

Figure 1: Location map showing calibration wells and PSDM area.

Density
2.1 2.2 2.3 2.4 2.5 2.6
9000

Sonic
2

170 150 130 110 90 70 50


9000

150

0
9000

100

50
9000

10000

10000

11000

11000

12000

12000

13000

13000

13000

14000

14000

14000

14000

15000

15000

15000

15000

16000

16000

16000

16000

10000

10000

11000

11000

12000

12000

13000

Figure 2: Well logs before (purple) and


after (blue) editing process.
Interval Velocity
5000
5000

7000

9000

11000

2 2.1 2.2 2.3 2.4 2.5 2.6 2.7

Figure 3: Typical well log suite generated from


editing process. Logs include Fluid Pressure,
Lithology Fraction, Sonic, and Density.
13000

15000

7000

Depth

9000

11000

13000

15000

17000

Figure 4: Well velocity (blue) compared


with extracted tomographic velocity (red)

5,000

10,000

15,000

20,000

Figure 5: Tomographic MVA volume

5,000

10,000

15,000

Productive Area

20,000

Figure 6: Pore-pressure volume.

Vous aimerez peut-être aussi