`
`
`Ex. PGS 1073
`EX. PGS 1073
`(EXCERPTED)
`(EXCERPTED)
`
`
`
`
`
`
`
`
`
`MAMDOUH R. GADALLAH
`
`EX. PGS 1073
`
`
`
`RESERVOIR SEISMOLOGY
`
`GEOPHYSICS IN NO NTECHNICAL LANG UAG E
`
`MAMDOUH R. GADALLAH
`
`pennWell Books
`
`PENNWELL PUBLISHING COMPANY
`TULSA, OKLAHOMA
`
`
`
`Copyright © 1994 by
`PennWell Publishing Company
`1421 South Sheridan/P.O. Box 1260
`Tulsa, Oklahoma 74101
`
`library of Congress Cataloging-in-Publication Data
`
`Gadallah, Mamdouh R.
`Reservoir seismology: geophysics in nontechnical language /
`Mamdouh R. Gadallah
`p. cm.
`Includes bibliographical references and index.
`ISBN 0-87814-411-0
`1. Seismic prospecting.
`TN269.8.G33 1994
`622'.1592—dc20
`
`I. Title.
`
`93-39214
`CIP
`
`All rights reserved. No part of this book may be reproduced,
`stored in a retrieval system, or transcribed in any form or by
`any means, electronic or mechanical, including photocopying
`and recording, without the prior written permission of the
`publisher.
`
`Printed in the United States of America
`
`1 2 3 4 5 98 97 96 95 94
`
`
`
`86 CHAPTER FIVE
`
`FIGURE 5-40. The stacking process
`
`*+•
`
`*+ = <>
`
`W {1) TWO PEAKS AT SAME ARRIVAL TIME
`
`(2) PEAK AND TROUGH AT SAME TIME
`AND WITH SAME AMPLITUDE
`(3) TROUGH AND PEAK DISPLACED
`LESS THAN ONE PEAK WIDTH
`
`(4) TWO PEAKS AT DIFFERENT TIMES
`
`DATA PROCESSING OBJECTIVES
`One of the most important objectives of seismic data processing is to
`produce readily interpretable sections. These sections must be geologi
`cally sound and represent the subsurface of the earth. They should be
`good enough to image the subsurface traps, where other seismic tech
`niques may be used for qualitative analysis of s ome petrophysical prop
`erties of the reservoir rocks. Seismic sections are used for special appli
`cations such as direct hydrocarbon indicators, porosity presence, and
`some other applications that will be reviewed later in the text.
`
`FACTORS AFFECTING SEISMIC DATA PROCESSING OBJECTIVES
`Seismic data processing objectives are strongly affected by field data-
`acquisition parameters. They are also affected by the quality of the field
`work and the condition of the equipment. There is an impression that
`field problems can be solved in the data processing stages, using the
`magic of the computer software and the muscle of the hardware. The bad
`news is that some (but not all) field problems can be solved in the data
`processing center.
`Marine conditions strongly affect the quality of the acquired data.
`Rough seas can cause streamer feathering, and a hard bottom will gen
`erate a variety of m ultiple reflections.
`Seismic data are usually collected under conditions that are less
`than ideal. Although we may not have the most desired information, data
`
`
`
`SEISMIC DATA PROCESSING 89
`
`Stage 3. Velocity analysis
`Velocity spectra—constant-velocity stack
`Design of front-end mute
`Stage 4. Velocity applications
`Normal moveout check
`Stage 5. Correlation statics
`Surface-consistent residual statics
`Application of statics to sort
`Re-run velocity analysis with refined static
`Refined surface-consistent residual statics
`Correlation statics
`Stage 6. Final stack
`Post-stack filter and decon tests
`Post-stack enhancement programs
`Migration
`Seismic inversion
`Special post-stack applications
`
`SUMMARY AND DISCUSSION
`There is a belief that the powerful computers and the sophisticated soft
`ware applications can solve field problems. The bad news is that the com
`puter software will not solve all the problems. If the problem of the data
`is resolution and insufficient data acquired in the field, or errors in the
`field parameters or malfunctions of the field equipment, computer soft
`ware cannot retrieve such information because it is not present in the
`field seismic data.
`Data processing is a very powerful tool in manipulating the field
`information. It converts the raw data with all information, including noise
`and distorted information, to meaningful seismic cross-sections that rep
`resent vertical slices through the subsurface. These slices represent geo
`logical information and potential hydrocarbon traps.
`We have learned a generalized seismic data-processing flow. Our in
`tention is to explain the physical meaning of some "buzz" words you hear,
`such as demultiplexing, which is rearranging the samples in the traces and
`across the records boundaries to make the data in trace sequential form
`for further software application. Cross-correlation is used to extract infor
`mation from seismic data acquired by the Vibroseis method; it is used to
`solve statics problems and for many more seismic data-processing appli
`cations. Gain is applied to compensate for spherical divergence and to
`preserve the relative true amplitude, which can be used as hydrocarbon
`indicator.
`Convolution is used to filter out some undesirable frequencies while
`deconvolution is a process to attenuate short-period multiple reflections
`
`
`
`3-D SEISMIC SURVEYS 221
`
`Swath shooting is economical. Better velocity analyses are obtained
`in data processing, since velocity analysis is reduced to one direction.
`Some advantages of swath shooting, if there are no environmental restric
`tions which will cause skips, are a uniform fold and good offsets. This
`type of shooting is best suited to simple structures and allows easy rect
`angular bins.
`
`3-D SWATH DESIGN
`3-D swath design is operationally simple over a large survey area, and it
`usually requires multiline equipment and a high number of channels.
`Each line of the swath is the same dimensions as in a 2-D spread. Thus,
`a five-line swath needs about five times the channel capacity of the con
`ventional 2-D line.
`Sources and receivers are one-directional. This is very important in
`calculating the angles and determining the offset to get better velocity
`analysis.
`
`BINS
`A bin consists of cells with dimensions one-half the receiver group spac
`ing in the in-line direction and nominal line spacing in the cross-line di
`rection (same as CDP in 2-D da ta processing). Traces that fail within a
`cell make up a common-cell gather. In marine 3-D surveys, not all of
`these traces are from the same shot line because of cab le feathering.
`Sorting of the data into cells is called binning. Bin size is a function
`of the velocity of the target, maximum frequency of the area, and dip
`angle of the target horizon. Figure 10-10 illustrates this relationship.
`Figure 10-11 illustrates the attribute analysis for each bin, which are:
`• Subsurface: target imaging.
`• Fold: its effect on S/N ratio and attenuating multiples.
`• Azimuth: which will effect the velocity analysis and give better imag
`ing for the dipping structure.
`• Offset range: will affect the velocity analysis to get the best stacking
`velocities and be able to get enough dynamic correction discrimination
`to attenuate multiples.
`
`FULL RA NGE 3-D
`A full azimuth range 3-D is well suited to small surveys, where source and
`receiver azimuths are multidirectional. Since the number of rollalong sta
`tion moves is large, this type of shooting is especially wellsuited for com
`plex structures. Because of the great volume of d ata collected, the pro
`cessing cost is fairly high.
`
`
`
`222 CHAPTER TEN
`
`FIGURE 10-10. Bins
`
`SAMPLING BASIC UNIT
`
`B I N
`
`S t H
`
`d a
`
`X
`
`S I N
`
`X
`
`>x
`
`B I N
`
`8 ( N
`
`X
`
`dG • GEOPHONE ARRAY INTERVAL
`<*S • SHOT INTERVAL
`B - V/{2*Fm*sin e) WHERE:
`B • BIN SIZE
`V - VELOCITY OF THE TARGET
`Fm • MAXIMUM FREQUENCY
`6 • DIP ANGLE OF THE TARGET
`
`FIGURE 1 0-11. Attribute analysis for each bin
`
`SUBSURFACE
`
`FOLD
`
`SUBSURFACE - TARGET
`
`FOLD - S/N, MULTIPLES
`
`AZIMUTHS - STRUCTURE. DIP
`
`OFFSET-VELOCITIES, MULTIPLES
`
`AZIMUTH
`
`O F FS E T RANGE
`
`
`
`3-D SEISMIC SURVEYS 223
`
`DESIGN OF MAXIMUM OFFSET
`The design of maximum offset, for both the nondipping and dipping case,
`is discussed in detail in Appendix B.
`
`3-D DATA PROCESSING
`LAND DATA PROCESSING
`
`Almost all concepts of 2-D seis mic data processing apply to 3-D data pro
`cessing, although some complications may arise in 3-D geometry, quali
`ty control, statics, velocity analysis, and migration. Editing of very noisy
`traces, spherical divergence correction for amplitude decay with depth
`and offset, deconvolution, trace balancing, and elevation statics are done
`in the preprocessing stage before collecting traces in common cell gath
`ers. Sorting in common cell gathers introduces a problem if there are
`dipping events, because there are azimuthal variations in the NMO appli
`cation within the cell.
`
`MARINE DATA PROCESSING
`After preprocessing, the data is ready for common-cell sorting, and a grid
`is superimposed on the survey area. This grid consists of c ells with di
`mensions one-half the receiver group spacing in the in-line direction and
`nominal line spacing in the cross-line direction. Traces that fall within the
`cell make up a common cell gather. Due to cable feathering, not all these
`traces are from the same shot line. The same data processing techniques
`in 2-D ma rine apply to 3-D marine data.
`
`APPLICATIONS OF 3-D SEISMIC DATA
`
`ENHANCED OIL RECOVERY
`A number of techniques will allow mapping of EOR p rocesses. One of the
`most familiar techniques uses 3-D s eismic, borehole seismic, and mi-
`croseismic. Defining the EOR front depends on the change of density of
`the reservoir rock as a result of the process. Commonly used techniques
`that can induce substantial density changes in the rock include steam
`flooding and in-situ combustion. The particular seismic method used
`depends on the goal to be accomplished. For example, a complete de
`scription of the developing front can be obtained by using 3-D seismic
`mapping, while limited knowledge can be obtained by using convention
`al 2-D seismic along the line.
`Several considerations are involved in field application of 3-D se is
`mic mapping in enhanced oil-recovery projects using steam injection.
`
`
`
`368 GLOSSARY
`
`Exploitation
`Hie development of petroleum reservoirs; wells are drilled to optimally
`drain the reservoir and have fairly low risk. The development wells may
`have several producing wells in the nearby drilling or spacing units.
`Extrapolation
`Projection or extension of unknown value from values within known ob
`servations or interval.
`'
`Fault
`Discontinuity in a rock type due to a break caused by tensional forces
`which cause normal faults or compressional forces which cause reverse
`faults.
`Fermafs principle
`See bow-tie effect (buried focus).
`First break times
`The first recorded signal from the energy source. These first breaks from
`reflection records are used to obtain information about the near-surface
`weathering layers. According to SEG polarity standard, an initial compres
`sion usually shows as a downkick.
`Flat spot
`A horizontal seismic reflection due to an interface between two fluids
`such as gas and water.
`Fold (Structure)
`An arch in a rock layer. A fold is usually formed as a result of def orma
`tion of rock layers by external forces. Folds include anticlines, cynclines,
`and overturns, etc.
`Fold (seismic)
`The multiplicity of common mid-point data.
`Formation
`A distinctive lithological unit or rock type.
`Formation velocity
`The speed by which a certain type of wave travels through a particular
`formation.
`Floating point (recording)
`A number expressed by the significant figures time a base raised to a
`power. This prevents the loss of a very small or a very large number.
`Computer uses bases that are a power of two rather than the base ten.
`Frequency
`A periodic waveform repetition rate in a second. Measured in cycle per
`second or Hertz.
`
`