`
`
`Ex. PGS 1081
`EX. PGS 1081
`(EXCERPTED)
`(EXCERPTED)
`
`
`
`
`
`

`

`Distinguished Instructor Series, No. 1
`
`Society of Exploration Geophysicists
`
`Time-Lapse Seismic
`in Reservoir Management
`
`Ian Jack
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`Copyright ¸ 1997
`Society of Exploration Geophysicists
`Box 702740
`Tulsa, OK USA 74170-2740
`
`
`
`Reprinted as Distinguished Instructor Series No. 1 in 1998
`
`All fights 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 mechanic!, induding photocopying and recoMing, without
`prior written permi(cid:127)ion of the publisher.
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`lan Jack, BP Exploration
`
`40%
`
`46%
`
`59%
`
`10%
`
`The first difference section, which is the 80-fold minus the 40-fold, is shown in Fig.
`4.B.6. The same plot gain is applied as for the benchmark and for all subsequent differ-
`ence plots. This table lists the displays and gives the rms amplitude of the difference
`section as a percentage of the benchmark section.
`ß Fig. 4.B.6(cid:127)80-fold minus 40-fold
`ß Fig. 4.B.7(cid:127)80-fold minus 20-fold
`ß Fig. 4.B.8(cid:127)80-fold minus 10-fold
`ß Fig. 4.B.9-(cid:127)4-km minus 3-km cable length
`ß Fig. 4.B.10--4-km minus 2-km cable length
`ß Fig. 4.B.11--4-km minus 0.5-km cable length
`ß Fig. 4.B.12(cid:127)10-fold
`ß Fig. 4.B. 13(cid:127)25-m cross-line spacing versus
`50-m cross-line spacing, interpolated
`
`40%
`
`72%
`
`60%
`
`84%
`
`The differences in each case are substantial, with the least being the loss of the outer
`1,000-m of cable. No cross-equalization has been attempted on these, and it is possible
`that doing so might just distort any differences that may be due to time-lapse effects
`such as reservoir depletion.
`Finally, two excellent examples of land 3D acquisition footprints caused by field
`geometry are shown in Fig. 4.B.14. Land data is sampled much less well than marine
`data and is more prone to these artifacts, which are difficult to remove in the differenc-
`ing operation that takes place in time-lapse work. In differencing operations, noise
`becomes much more of an issue!
`
`4C. Some positioning and timing repeatability issues and warnings
`This section expands on some of the items listed in section 2C, and imparts a healthy
`suspicion about the effects of positioning and timing accuracy The data sets analyzed
`are all marine, but references elsewhere suggest that land data repeatability is at least as
`troublesome. In differencing operations, noise becomes much more of an issue!
`
`This section includes the following:
`ß Controlled tests of position degradation, using the computer to misposition data.
`ß Field acquisition trials with state-of-the art positioning, repeated three days apart.
`ß An analysis of some recent repeat surveys, 1995 and 1996.
`ß An analysis of some legacy data, 1985 compared with 1995.
`ß A brief review of the literature.
`
`Controlled tests of position degradation: A swathe of data has been taken from a 3D
`marine data set and processed to stack. It was not migrated, as some of the systematic
`effects resulting from the deliberate degradation would be obscured. The benchmark
`line is in Fig. 4. C. 1.
`
`Distinguished Instructor Short Course ß 4-3
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`Time-Lapse Seismic in Reservoir Management
`
`A constant shift of 25 in in the cross-line direction, applied to the antenna position
`of every fourth sail line, then differencing the section from the benchmark, gives the
`result in Fig. 4.C.2. This section contains 47% of the rms amplitude of the benchmark
`and represents mainly the effect of the changing geology for a 25-m constant position
`error. Constant shifts gave higher residuals than variable shifts that were also applied
`(hardly surprising). Three of these variable shifts are shown in the following figures.
`ß Fig. 4.C.3 (25%) shot location error, sinusoidal, with peak amplitude 12.5 in in-line.
`ß Fig. 4.C.4 (16%) shot location error, sinusoidal, with peak amplitude 12.5 in X-line.
`ß Fig. 4.C.5 (24%) shot location error, sinusoidal, with peak amplitude 25.0 rn X-line.
`The in-line residuals are greater due to the effect of incorrect moveout resulting from
`the shot position error.
`Not everyone has been confident of streamer positioning accuracy, especially on
`some older surveys, so errors were deliberately introduced to quantify these effects. In
`these tests the streamer was rotated on each fourth sail line by mispositioning the tail-
`buoy by several amounts, both fixed and variable. Three of the results are shown in the
`following figures.
`ß Fig. 4.C.6 (20%) streamer location error, constant, by 40-m tailbuoy rotation.
`ß Fig. 4.C.7 (32%) streamer location error, constant, by 80-m tailbuoy rotation.
`ß Fig. 4.C.8 (19%) streamer location error, variable, by 80-m peak tailbuoy rotation.
`Streamer length is 4,000 m.
`These tests show that we can probably expect to find considerable residuals when
`we difference data sets on which positioning errors were of a size considered to be com-
`mon on surveys prior to the early 1990s. The results shown will be "best case," as no
`geometry or other acquisition differences exist, except those applied in what was essen-
`tially a data processing exercise.
`Field acquisition trials: The second set of examples should also be close to a "best
`case" since it uses hydrophone detectors ploughed into the seabed, in the joint
`
`BP/Shell/Geco-Prakla time-lapse experiment in Foinaven, West Shedands. The second
`acquisition at this location was repeated with the same vessel and air gun system just
`three days after the first. Fig. 4.C.9 shows the field layout for the test.
`A set of data from two corresponding shot-points with identical "true amplitude
`recovery" scaling, and their subtraction, is shown in Fig. 4.C.10. Rather surprisingly,
`the subtraction contains 35% rms of the amplitude of the input data. Limiting the band-
`width reduces this, but only slightly, to 26%. Stacking the data reduces the figures to
`27% and 23%, respectively, Fig. 4.C.11. Prestack matching of the data reduced the dif-
`ferences very slightly (about 1%), then prestack plus poststack matching by a further 2%.
`These results triggered an investigation into the reasons for their magnitude, which
`involved examining positioning, water velocity stability, and timing. The first breaks
`from corresponding shots were flattened with appropriate moveout, and travel time dif-
`ferences plotted against shot-receiver distance. The gradient of this plot can explain
`either a water velocity change of 2.5 m/s or a shot positioning difference of 0.7 m.
`Equipment stability was assessed by picking first break timing differences by cross-
`correlation, from shots vertically above receivers. A scatter of up to a millisecond was
`
`4-4 ß Society of Exploration Geophysicists
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`Time-Lapse Seismic in Reservoir Management
`
`Figure 4. C. 1
`
`Navigation Benchmark
`
`!%'
`
`Figure 4.C.2
`
`Antenna Shift 25m cross-line, Difference
`
`F(cid:127)4. C2
`
`Figure 4.C.3
`
`Shot Location, in-line variable 12.5m, Difference
`
`4
`
`4-16 ß Society of Exploration Geophysicists
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`/an Jack, BP Exploration
`
`Figure 4.C.4
`
`Shot Location, cross-line variable 12 5m, Difference
`
`Figure 4.C.5
`
`Shot Location, cruse-line variable 25m, Difference
`
`..
`
`(cid:127)
`
`(cid:127) 'f'
`
`?
`
`.'1 (cid:127)
`
`'(cid:127)'
`
`'I!.
`
`. 't,
`
`Figure 4.C.6
`
`Streamer Location, constant 4Ore, Difference
`
`
`
`
`
`'(cid:127)'_=_..(cid:127)___(cid:127)_. . (cid:127)'. ",' (cid:127)",' '(cid:127)' . . (cid:127)" ...... '1"
`
`Distinguished Instructor Short Course ß 4-17
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`Time-Lapse Seismic in Reservoir Management
`
`Figure 4.C. 7
`
`Sbeamer Location, constant 8Ore, Difference
`
`Figure 4.C.8
`
`Figure 4.C.9
`
`Femaven Repealabiay Test
`
`4-18 ß Society of Exploration Geophysicists
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`Figure 4.C.10
`
`Figure 4.C. 11
`
`Figure 4. C. 12
`
`lan Jack, BP Exploration
`
`Fi(cid:127)4.C.10
`
`Distinguished Instructor Short Course ß 4-19
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`

`

`Time-Lapse Seismic in Reservoir Management
`
`Figure 4.C.13
`
`Survey Geometries 1985,1995,1996
`
`Figure 4.C.14
`
`Figure 4.C. 15
`
`CompnFJfon of 1995 and 1996 raw shot records
`
`4-20 ß Society of Exploration Geophysicists
`
`Fio,4.C.'13
`
`[:1(cid:127) 4.C 14
`
`I=ig.4.C.15
`
`Downloaded 08/27/14 to 173.226.64.254. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/
`
`