moscow

Oil Recovery 2003 – 1st
International Conference and Exhibition – Modern Challenges in Oil recovery
FROM ELASTIC INVERSION DATA TO RESERVOIR MODEL
- 4D EXPECTATIONS AND HISTORY MATCHING
Stig-Arne Kristoffersen 1
, Dave Davies 2
, Klaus Bolding Rasmussen 1
1
Ødegaard AS 2
Ødegaard UK Ltd
(e-mail: sak@oedegaard.com)
Abstract
This paper discusses the utilization of Inversion data in a IOR process of a field
offshore North Sea.
Important elements in this paper is where inversion data can be used, how they
are implemented in the IOR process and when maximum effect of data can be
expected.
Methodology on how to intregrate reservoir model physical properties and
seismic acoustic properties through rock physics understanding is presented in
this paper.
New methodologies are developed in order to utilize the reservoir simulator grip
physical properties in order to use the data on seismic workstations.
Methodology to close the circle and go from seismic acoustic properties and
populate the reservoir simulator is the ultimate goal in this process.
Introduction
Input of AVO seismic data provides us with Acoustic Impedance and Poission’s
ratio resultant data, which we can use for lithology –fluid predictions. This is
enabled through integration of acoustic well data and acoustic seismic data,
figure 1.
The goal is to make a final product, which is reservoir unit bodies, which can be
compared to reservoir simulation description, figure 2.
Ødegaard AS has a lithology cube, which describes the probabilities for fluid and
litholgies within the seismic volume.
The benefits of the lithology cube are that it provides us with direct sub-surface
measurement of rock and fluid properties such as;
N/G
Porosity
Fluid type
Saturation
Litholgy (shale, sand, hydrocarbon sand, brine sand, limestone, dolomites etc)
The lithology cube also provides us with direct measurement of time-varying
rock and fluid properties in the sub-surface, such as;
Saturation changes
Pressure changes
1

Temperature changes
Rock body changes
The workflow of lithology cube generation is ;
Generation of Absolute Acoustic Impedance volumes (AAI)
Generation of Poisson’s ratio cube, (PR)
From well logs – base line lithology and fluid information is retrieved.
Then we define the lithological probabilities, forward and inverse from cross-
plotting AAI and PR values in cube.
Next step is to make a Time-lapse lithology cube estimation, and then make
difference cubes – a 4D cube.
Inversion of seismic amplitude data will enhance resolution of seismic data, and
dampen the noise in dataset. In addition we will be able to estimate physical
properties.
Ødegaard AS – ISIS AVO Inversion has one big advantage, and that is its
potential to inverse large volumes of 2D and 3D data.
Output from inversion of seismic data is Absolute Acoustic Impedance; shear
impedance and Poisson’s ratio or other acoustic parameters required by the user.
Determination of the fluid and lithology separation in both well and seismic data
is crucial in studies of this format, figure 2.
It is therefore required to have a balanced and well-controlled inversion process,
which do describe the actual seismic data, and is by no means steered by too
many models in your prior model.
Ødegaard AS philosophy is to let the seismic data tell as much as possible about
the high-frequency content in data, and our low frequency model is only guiding
the absolute levels in the very low frequency domain, normally not present in the
seismic data.
The angle stack inversion process should therefore be robust enough to describe
the absolute best sub-surface model, and this can be achieved through global
optimalisation of the data.
The integration of seismic and well data together with any interpretation
performed in an area is illustrated best through a workflow diagram as seen in
figure 3 a and 3b.
Here we can see the way seismic angle-stacks are put into work with wavelet
estimation and through log calibration the well data is incorporated. The Low
Frequency (LF) model is implemented into the process simultaneous to the
inversion process, in order to prohibit any notches or contamination in the
seismic data.
2

ISIS inversion are run simultaneously on all angle stacks, which allows us to
derive AAI and PR together with density volumes, for usage into the lithology
and fluid predictions.
Each angle stack has their individual wavelets, which are compensated for
frequency, energy and phase variation.
This provides us with robust inversion results, which will be used to describe
physical parameters from the acoustic domain, figure 4.
The lithology-fluid prediction is then used for a probability distribution in the
seismic volume. This is closely calibrated to the wells, which enables us to QC
the results obtained in this process, figure 5.
The final product applying reservoir properties either existing in reservoir model
or as seen in wells, will provide us with a reservoir body, which can be used for
hydrocarbon volumetrics, figure 6.
Ødegaard AS has performed about 25 4D projects worldwide, and the experience
gained in these projects tells us that the time and depth domain has to be closer
connected than seen today. This means that reservoir simulation grids have to
be visualized and compared with seismic inversion data. Comparison of these
two data sets, enables the user to explain the robustness of the reservoir model
as well as utilizing the seismic capabilities. This enables us to make a inter-well
trends and utilize well information integrated into seismic as hard fact points,
figure 7.
Legacy data can be used in a basic loop of workflow, where reservoir model data
is integrated into seismic inversion data, and compared for body identification.
Ødegaard AS has in several projects enabled the client to make a 4D loop where
simulation data is integrated into seismic inversion data, and difference cubes
are made for testing of reservoir model validity in a field.
This again is compared towards the geological model of the field, figure 8.
Another area of usage is in feasibility studies in fields, as to when to aquire new
seismic data, and which expectancies of changes can we have when it comes to
AAI and PR as well as density and shear impedances in a field. These acoustic
parameters are modeled from the reservoir model parameters such as porosity,
saturation and pressure changes over time.
Ødegaard AS has used this methodology both on clastic as well as carbonate
rocks, and found it to be very useful for seismic acquisition planning purposes.
In the North Sea, Ødegaard AS has performed in excess of twenty 4D and IOR
projects, including Nelson, Tern, Hudson, Ninian, Ekofisk, Gullfaks, Oseberg
South, Njord, Grane and Ormen Lange are some of these fields where 4D
inversion seismic has proven to be vital for reservoir management and IOR
steering of the field.
3

Conclusion
By bringing together forward modeling from reservoir Simulator with time-lapse/
4D observations, 4D AVO inversion can contribute significantly to reservoir
modeling, IOR steering of a field.
To obtain optimal reservoir management, the reservoir models should be used to
secure optimal reservoir management both on the short- and on the long term.
Real time reservoir management has increased its importance, therefore
Increased reservoir knowledge (like through 4D inversion seismic) is used more
actively to locate infill-wells and to adjust and control the recovery schedule. This
combined with the possibility of using combinations of several different IOR
methods simultaneously and monitor the effect on these methods in 4D inversion
seismic data Is important.
IOR is a true multidisciplinary exercise, where inversion data plays as equally
important role as the well control in a field.
Experience tells us that increased reservoir knowledge, is one of the most central
categories of IOR management control and is vital in order to optimize and
verify the other categories in IOR management.
We therefore state that inversion data is no longer a “nice to have” feature, but
must be treated as “need to have” product in any IOR management control
system one use in a field.
Focus on enhanced oil and gas recovery in a field comes through focused R&D
and technology watch which enables integration of data types available, and
utilization of data availability.
Increased oil recovery can only be obtained through focus on integrated
discipline areas. From a G&G standpoint, it is important to contribute to link the
reservoir engineering and engineering disciplines with the G&G disciplines.
This can be achieved through more focus on usage of 4D seismic into geological
model side, which again can be translated into reservoir model tools.
Areas to focus usage of 4D inversion seismic are in geosteering, visualizations,
monitoring of liquid and gas reservoir movements
Combined with usage of new monitoring techniques and utilizing the 4D
inversion data for optimal reservoir management and transfer it into real time
reservoir management systems.
4D inversion seismic can effectively be used to reduce residual oil saturation and
improve sweep in fields.
4

“Integrated volume interpretation of 4D inversion difference (oil water contact
movement) and oil sand probability show areas of unswept oil, highlighting infill
opportunities. Early results from infill drilling have validated the method realising
the potential economic benefits of 4D seismic technologies.”
A.T. McInally, Shell Exploration and Production U.K.et.al, Petroleum geoscience
February, 2003.
Figure List;
Figure 1;
Figure 2;
5

Figure 3a;
Figure 3b;
6

Figure 4;
7

Figure 5;
Figure 6;
8

Figure 7;
Figure 8;
9

10

moscow

Related slideshows

More Related Content

moscow