La lecture en ligne est gratuite
Le téléchargement nécessite un accès à la bibliothèque YouScribe
Tout savoir sur nos offres
Télécharger Lire

Brief Introduction to Database Concepts

46 pages
  • cours - matière potentielle : that a student
  • mémoire
  • fiche de synthèse - matière potentielle : the entities
Brief Introduction to Database Concepts Andrea Rodrıguez Summer School - Castellon 2004 Department of Computer Science University of Concepcion, Chile 1 Introduction Information and data are different. Information is understood by a person. Data are values stored on a passive medium like a computer disk. The purpose of a database management system (DBMS) is to bridge the gap between information and data - the data stored in memory or on disk must be converted to usable information.
  • connection between entities
  • rest of the system
  • entity
  • section
  • model
  • attribute
  • relationships
  • schema
  • class
  • database
Voir plus Voir moins

Model-based Inversion
Challenges the
Technology Transfer
• AUGUST 1-5, 1999 •
Editors: F. Gasparotto, K.-M. Strack, S. TreitelORGANIZING COMMITTEE
Chairman: K.-M. Strack, KMS Technologies
Technical Program: M.K. Sengupta, Schlumberger Geco-Prakla
Abstract Volume: F. Gasparotto, Exxon Production Research Co.
K.-M. Strack, KMS Technologies
S. Treitel, TriDekon Inc.
Organizing Committee: J. Berryman, Lawrence Livermore NL;
N. Dutta, Baker Hughes
K. Vozoff, HarbourDom
A. Weglein, ARCO
SEG Research Committee: L. Thomsen, BP - AMOCO
When it was understood that geophysical data could help find petroleum, quickly the
quest for the understanding of the ‘physics’ appeared. The mind of the geophysicist came up
almost simultaneously with a model of the Earth that explained the behavior of the
measurements. The models were initially even cruder than the measurements but quickly they
were refined from simple models to more complex ones. Limiting was then the computer
power and geophysicists were always lagging behind geologists. Today, the model complexity
is almost where the geologic models are. Immediately, the exploration managers want to know
about risks and reliability of the results. This is still a point where the geophysicists – the one
that uses inversion techniques- will answer " What do you want it to be?" In reality, out tools
give us all the answers if we just understand it.
The desire to invert geophysical recordings continues to be among our fondest dreams.
While a complete, unambiguous and unique picture of the subsurface must forever remain
beyond our grasp, the past years have seen significant progress in this exciting area of our
profession. While a significant body of literature dealing with inverse theory has existed for
many decades, real-life practical applications in exploration geophysics have been quite slow
to emerge. One reason for this might well be that the wall between those solving erudite
equations and those seeking practical applications continues to be thick and high. But clearly
our own discipline of exploration geophysics cannot profit from all the exciting theoretical
accomplishments unless the means can be found to transfer them to the user. From the outside
it often looks as if the products – namely inversion software and services – are not mature. It
is precisely for this reason that our workshop will emphasize the technology transfer aspects of
geophysical inversion and nurture a better understanding of the limitations.
The response to our call for papers and posters has been very gratifying, particularly if we
bear in mind that our industry is going through such difficult times. The workshop program
contains a wide variety of contributions, which demonstrate quite clearly that geophysical
inversion is beginning to make a practical impact. The formal talks, the posters, the panel
discussions and perhaps most importantly, the informal technical exchanges that always
develop during small technical gatherings will hopefully do much to advance this exciting
branch of our profession.
Here we need to say something about the future without having a crystal ball. So let’s
simply say: If we don’t know about the initial part already today it will not be part of our
everyday future 10 years from now. Based on that the future will clearly complete the
integration into the 3D volume using seismic data and other data as constraints. Inversion
methods will be the key tool to save time and analyze risk factors. The 3D volume integration
(and the first example are already there), we will then improve the model descriptions and get
the models more analytical. This will then even more improve the signal-to-noise in model
interpretation and subsequently yield more reserves.
Fernanda Gasparotto, Mirnal Sengupta, Kurt Strack, Sven Treitel➤

Organizing Committee..................................................................1
Welcome ......................................................................................3
Technical Program ........................................................................7
Abstracts ....................................................................................11
List of participants ......................................................................45TECHNICAL PROGRAM
8:00 a.m. – Noon Inversion Methods & Overviews I: General
Noon Lunch – Outside Patio
1:30 p.m. – 4:00 p.m. Poster Presentations: Methods & Case Studies
4:00 p.m. – 5:30 p.m. Panel Discussions: Inversion Non-Uniqueness
Dinner and time on your own
8:00 a.m. – Noon Inversion Methods & Overviews II: Well logs
Noon Lunch – Outside Patio
1:30 p.m. – 4:00 p.m. Poster Presentations: Methods & Well-log Applications
4:00 p.m. – 5:30 p.m. Panel Discussions: Inversion Technology Transfer
Dinner and time on your own
8:00 a.m. – Noon Inversion Methods & Overviews III: Seismic
Noon Lunch – Outside Patio
1:30 p.m. – 4:30 p.m. Poster Presentations: Methods & Seismic Applications
4:00 p.m. – 5:30 p.m. Panel Discussions: Enhancing Inversion Bandwidth & Accuracy
6:00 p.m. – 9:00 p.m. Reception and Dinner – Outside Patio
8:00 a.m. – Noon Wrap Up: The Road Ahead
Noon Workshop adjournedMONDAY, AUGUST 2
7:00 a.m. Breakfast – Los Vaqueros
8:00 a.m. Inversion Methods and Overviews I: General
Chairperson: Kurt Strack
8:15– 10: 15 a.m. Talks: 30 minutes each
Brian H. Russell: Technology Transfer of Inversion Methods
Chuck H. Wagner: Integration of Seismic Inversion into Development Drilling Process
Carlos Torres-Verdin: Geostatistical Inversion
M.K. Sengupta: Taking Advantage of Non-Uniqueness in Inversion
10:15 – 10:30 a.m. Coffee Break
10:30 – 12:00 p.m. Introduction to Posters: 10 minutes each
LUNCH BREAK – Outside Patio
1:30 p.m. – 4:00 p.m. Poster Presentations: Methods & Case Studies
Each poster board is 8 feet tall and 4 feet wide. However, the usable space is 7 feet 7 inches tall & 3 feet 8
1/2 inches wide. A presentation booth would have 2 of these poster boards.
3D velocity-depth model building using surface seismic and the wells: Paul Sexton, Paul
Williamson, & Phillippe Berther
Calibration of prestack inversion using Bayes’ theorem & well data: Phil D. Anno, Robert T.
Baumel, & Javaid A. Durrani
Integration of seismic inversion into the development drilling process:
Charles H. Wagner III, Edgardo L. Nebrija, Harold L. Triebwasser, Arthur E. Gregory,
Muhammad Saggaf, & Jung J. Kim
Seismic inversion of a reef complex in the South China Sea: Claire Sullivan, Chip Story, Patrick
Peng, Christoph Heubeck, & Lin Jian Dong
The 4-D Microgravity Method for Waterflood Surveillance: A Model Study for the Prudhoe Bay
Reservoir, Alaska: Jennifer L. Hare, John F. Ferguson, Carlos L. V. Aiken, & Jerry L. Brady
New method for 3D automatic volumic reconstruction, Validation through Geophysical
simulations: G. Courrioux, C. Truffert, S. Nullans, & A. Bitri.
New method to determine the lithology and the geometry of geological object in-depth: litho-
inversion 3D: Antonio Guillen, Miguel Bosch, V. Delos, & P. Ledru
Geostatistical Inversion and lithology discrimination: Honoring 3D seismic data and well logs to
delineate Sands thinner than tuning resolution: Carlos Torres-Verdin, German Merletti, Marcos
Victoria, Pieter van dar Made, & Paul van Riel
The role of a priori’ information and parameterization in real word application of the inversion of
potential field data: F. Boschetti, P. Hornby, F. G. Horowitz, & N. Archibald
4:00 p.m. – 5:30 p.m. Panel Discussions: Inversion Non-Uniqueness
Panelists: Daniel Hampson, M. K. Sengupta, Kurt Strack,
Carlos Torres-Verdin, Chuck Wagner
“Moderator” is underlined.
Dinner and time on your ownTUESDAY, AUGUST 3
7:00 a.m. Breakfast – Los Vaqueros
8:00 a.m. Inversion Methods and Overviews II: Well logs
Chairperson: Fernanda Gasparotto
8:15 – 10:15 a.m. Talks: 30 minutes each
Kurt Strack: Well-log Inversion Review
Doug Oldenburg: Inversion of potential and EM field data
Lee Bell & Larry Morley: Applications of Diving-Ray Tomography in Exploration
Victor Pereyra: Model Based 3D Travel Time Inversion
10:15 – 10:30 a.m. Coffee Break
10:30 – 12:00 p.m. Introduction to Posters: 10 minutes each
LUNCH BREAK – Outside Patio
1:30 p.m. – 4:00 p.m. Poster Presentations: Methods & Well-log Applications
Limits and possibilities using inversion with array Resistivity logging tools: Kurt Strack, Raghu
Chundru, Mikael A. Frenkel, Alberto G. Mezzatesta, & Z. Zhang
HDIL post-inversion processing: Achieving consistency with other logs: Mikael B. Rabinovich, &
Leonty A. Tabarovsky
Computational Intelligence Techniques for TEM Inversion: Hesham El-Kaliouby, & Mary M.
Inversion of cross-well EM data: Ping Zhang, & Mike Wilt
Comparison of model-driven, L1 Norm (Sparse-spike), and Recursive Inversion Methods:
Xin-Quan Ma, & Adrian Pelham
Lithologic tomography: method and 2D application: Miguel Bosch, Antonio Guillen, &
Patrick Ledru
Model Based 3D Travel Time Inversion: Victor Pereyra
Inverting the recording footprint by migration deconvolution: Jiansing Hu, Gerard Schuster, &
Paul Valasek
3D Prestack Kirchhoff Beam Migration for Depth Imaging: Theory and Data Examples: Y. Sun,
F. Qin, S. Checkles, & J. Leveille
4:00 p.m. – 5:30 p.m. Panel Discussions: Inversion Technology Transfer
Panelists: Fernanda Gasparotto, Mike Payne,
Brian Russell, Kurt Strack
Dinner and time on your ownWEDNESDAY, AUGUST 4
7:00 a.m. Breakfast – Los Vaqueros
8:00 a.m. Inversion Methods and Overviews III: Seismic
Chairperson: M. K. Sengupta
8:15 – 10:15 Talks: 30 minutes each
Daniel Hampson: Multi-attribute Inversion
Subhasis Mallick: Genetic Inversion
David Gray: Inversion for Rock Properties
10:15 – 10:30 Coffee Break
10:30 – 12:00 Introduction to Posters: 10 minutes each
LUNCH BREAK – Outside Patio
1:30 p.m. – 4:30 p.m. Poster Presentations: Methods & Seismic Applications
AVO inversion using multiple seismic attributes: Daniel Hampson, & Todor Todorov
Hybrid Seismic Inversion: A reconnaissance exploration Tool: Subhasis Mallick, & Jeffrey Lauve
Inversion for fundamental rock properties: F. David Gray
Inversion of Multi-Component Data to Improve Reservoir Description: Side Jin
Investigating the impact of anelasticity on model based AVO inversion: Arnim B. Haase, & William
N. Goodway
Inversion of offset stacks for discrimination of lithology: John V. Pendrel, Robert R. Stewart,
Jocelyn Dufour, Bill Goodway, & Paul van Riel
Inversion of seismic refraction data with unique averages a more robust alternative to the model-
based methods: Derecke Palmer
Inversion: Seismic feature versus Seismic Attributes: Fred Aminzadeh, & Quincy Chen
Cape Ford - 1, An Example of the consequences of true amplitude processing for inversion to
acoustic impedance and porosity prediction: Marianne Rauch, & Phil Woods
A Few Examples of Seismic Acoustic Impedance Inversion: Steve Clawson
4:00 p.m. – 5:30 p.m. Panel Discussions: Enhancing Inversion Bandwidth and Accuracy
Panelists: Doug Oldenburg, Chris Chapman, Subhasis Mallick,
Victor Pereyra, Kurt Strack
6:00 p.m. – 9:00 p.m. Reception and Dinner – Outside Patio
7:00 a.m. Breakfast – Los Vaqueros
8:00 a.m. – Noon Wrap Up: The Road Ahead
Open floor and panel discussion
Chairperson: Kurt Strack
Panel: Chairpersons from sessions & panels
Noon Workshop adjournedABSTRACTS
Technology Transfer of Inversion Methods
Brian Russell
SEG President
1998 – 1999
thInversion theory can be dated back to the early 19 century German mathematician
Carl Friedrich Gauss, who labored without the benefit of computers. Since Gauss’
time, and thanks to the introduction of computers, we have made a lot of progress
with respect to inversion theory and its application, and late twentieth century
theoretical geophysicists have applied it to almost every possible geophysical
problem. Despite the wide variety of applications available, inversion theory is still
not fully understood or accepted by a large number of practicing geophysicists,
especially those who make their living by exploring for, and developing, petroleum
reservoirs. Often, inversion is seen as something that is esoteric and nonessential,
and certainly not part of mainstream exploration.
Over the last twenty-five years, a number of attempts have been made to introduce
inversion technology to the practicing geophysicist, in the form of either consulting
services or software. These efforts have met with varying amounts of success,
depending on factors such as the complexity of the method, its robustness, its
success in detailing stratigraphic anomalies, and so on. Recently, there appears to
be a higher level of acceptance of inversion methods by the geophysical
community. But much work still needs to be done on the part of inversion
specialists in convincing their colleagues of the applicability of their methods to
practical exploration and development problems.
This paper will look at the development of inversion software from the mid-
seventies to the present, focusing on two techniques in particular:
! Post-stack impedance inversion.
! Pre-stack estimation of elastic parameters.
The above topics have been chosen from the wide array of inversion tools that are
available for two reasons. First, these are the methods that are most generally
associated with the inversion of seismic data and, second, the author has had a
personal involvement with the evolution of these methods in particular.
In the paper, I will focus on the necessary ingredients for the successful technology
transfer of inversion software. These include the factors discussed above, as well
as external factors, such as the development of interactive graphics, personal
workstations, faster CPU’s and larger file storage capability. I will also look at the
advantages of using inversion techniques in a typical exploration and production
cycle, as well as some of the pitfalls. Finally, I will make some predictions about
where I see inversion, and inversion software, heading in the next century or, dare I
say, millennium. Not having a crystal ball, I will undoubtedly be wrong! Geostatistical inversion and lithology discrimination:
honoring 3-D seismic data and well logs to delineate sands thinner
than tuning resolution
Carlos Torres-Verdín, Germán Merletti, and Marcos Victoria, YPF, S.A.
John Pendrel and Paul van Riel, Jason Geosystems
Synthesizing the information content available from both poststack 3-D
seismic data and well logs is one of the most common problems encountered in the
geophysical evaluation of hydrocarbon reservoirs. Borehole data possess superior
vertical resolution than seismic data. However, even in cases where wells are
closely spaced, their lateral resolution is seldom superior to that of 3-D seismic
data. It then becomes natural to advocate a procedure whereby one could
consistently integrate the best of both worlds, i.e., the high vertical resolution of
wellbore measurements and the high lateral resolution of 3-D seismic data. There
are fundamental problems, however, that need to be addressed before embarking on
one such procedure. For instance, the length of penetration and spatial resolution of
wellbore measurements do not overlap with those of surface seismic data. The
difference in spatial resolution and length of penetration can make the underlying
physics of the two measurements completely different in light of, for instance,
severe dispersion effects. Attribute mapping via cross-plots is a popular method
used to characterize the petrophysical nature to the seismic amplitude variations.
A common example is the mapping of porosity into the seismic volume.
Despite being appealing for its simplicity, this procedure can fail because of three
drawbacks: (1) wavelet superposition effects (tuning), (2) well-log measurements
are not representative of the physics governing the seismic data, and (3) the
relationship between the petrophysical variable (e.g. porosity) and the seismic
amplitudes is neither linear nor unique. In this paper, we describe a geostatistical
inversion procedure designed to effectively address these 3 problems in producing
3-D estimates of porosity within the seismic cube. Firstly, wavelet superposition
effects are reduced by estimating a seismic wavelet at key control wells and then
using this wavelet to invert the seismic data into acoustic impedances (AI).
Secondly, we make an attempt at putting the inverted AI´s on equal footing with
the wellbore AI´s by low-pass filtering the latter to have the same frequency
bandwidth of the seismic data. This exercise yields cross-plots of AI with respect to
porosity that can be used to characterize the petrophysical nature of the inverted
AI´s. Such cross-plots are treated not as linear maps but rather as cloud transforms
that in general non-uniquely relate a given value of AI with porosity (shaly sands
are but one example when such behavior may occur). However, if the vertical
resolution of the inverted AI´s is not high enough to discern individual sand bodies,
we resort to geostatistical inversion to extrapolate the wellbore data into the
reservoir but only to the extent allowed by the seismic data themselves. This is
done with a stochastic co-simulation technique that enforces variograms and
histograms sampled from the inverted AI´s and the wellbore data. The sampled
histograms are used to construct probability density functions from which the
random samples of porosity are drawn. These samples are then fed to the assumed
cloud transform to yield their corresponding AI values. Acoustic impedances
derived in this manner are subsequently transformed into a time series of reflection
coefficients, which are convolved with the seismic wavelet to simulate the seismic