See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/275465886

CALIBRATION OF SONIC LOGS FOR SEISMIC APPLICATIONS IN UPPER ASSAM

Conference Paper · January 2004

CITATIONS READS
0 3,574

4 authors, including:

Lalaji Yadav
Reliance Industries Limited
41 PUBLICATIONS   50 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Prediction of Archie parameters from grain size variations View project

All content following this page was uploaded by Lalaji Yadav on 27 April 2015.

The user has requested enhancement of the downloaded file.

 SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

CALIBRATION OF SONIC LOGS FOR SEISMIC APPLICATIONS IN UPPER ASSAM

Lalaji Yadav 1, Dilip Ghosh 2, S.N. Maurya 3 and A. N. Bhattacharya 1

1 Well Logging Services, ONGC, Nazira, Assam, India-785685

2 Sub Surface Team, Assam Asset, ONGC, Nazira, Assam, India-785685
3 A & AA Basin, Forward Base, ONGC, Nazira, Assam, India-785685

ABSTRACT

Merging of data from seismic surveys and well logs in geophysical exploration is a part of synergistic
approach for better understanding of geological model of the field. Sonic log provides continuous
records of different formation velocities especially at shallow and intermediate depths, which may not
be interesting from reservoir point of view. Exact knowledge of velocities is used for time to depth
conversion, computation of synthetic seismic sections, migration and processing of geologically
complex areas. Once hydrocarbon trap is proved by a discovery well, the whole seismic interpretation
is to be re-looked for a new range of possibilities, resolution and detailed study. Delineation of known
producing formations then becomes more discernible on seismic sections studied in conjunction with
geophysical borehole measurements.

While using sonic logs for seismic purposes, one has to adjust the sonic logs according to seismic
measurements in the boreholes because sonic times obtained through the integration of sonic logs
usually differ from those obtained by means of a well seismic. The reasons for drift range from basic
discrepancies between two approaches due to different geometry / frequency of measurement
principles etc. to more trivial disturbances in the formation i.e. alteration / invasion, which affects the
sonic logs resulting in cycle skipping, detection of mud arrivals in large holes. Since the check shot /
VSP method is simpler and corroborates with reflection survey results, sonic is more susceptive.
Therefore, it is necessary to calibrate the sonic logs in order to eliminate these possible errors to use it
for seismic application for a particular area.

In the present study interval velocity, average formation velocity are computed from both one way
seismic time and integrated sonic time data available over sixteen wells spread all over Upper Assam
oil fields shows that the negative drifts for the calibration of sonic transit time. Amount of drifts is more
in Girujan Clays about more than half of the total drift. Interval velocity values derived from seismic
one way time are showing finer variations, which are some times cyclic in nature i.e. higher velocity
followed by lower one, as compared to interval velocity values derived from sonic. These variations will
result in creating fictitious reflections on synthetic seismograms, which are not seen on log
measurements. Block shifting of sonic log ensures integrated travel time at par with seismic one-way
time by preserving formation signatures required for seismofacies mapping. It is worth mentioning that
sonic log has not lost even a bit of its grace as continuous velocity log even with recording of well
seismic data / VSP.

This method of correcting sonic logs for seismic applications will certainly provide a value addition for
the realistic interpretation of seismic data in future study of Upper Assam fields and the method may
be extended for other area too.

KEY WORDS :: Sonic Logs Calibration, Drift Analysis, Upper Assam

1
 SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

INTRODUCTION

Merging of data from seismic surveys and well logs in geophysical exploration is a part of synergistic
approach for better understanding of geological model of the field. Sonic log provides continuous
records of different formation velocities especially at shallow and intermediate depths, which may not
be interesting from reservoir point of view. An exact knowledge of velocities is used for time to depth
conversion, computation of synthetic seismic sections, migration and processing of geologically
complex areas. As such velocity is an extremely important parameter for seismic interpretation. Once
hydrocarbon trap is proved by a discovery well, the whole seismic interpretation is to be re-looked for a
new range of possibilities, resolution and detailed study. Delineation of known producing formations
then becomes more discernible on seismic sections studied in conjunction with geophysical borehole
measurements. One feels almost apologetic to mention this, but there have been cases where dry
holes have been drilled because no one has gone back to look at seismic sections where producing
sand was as plain as pikestaff (Anstey, 1977).

While using sonic logs for seismic purposes, one has to adjust sonic logs according to seismic
measurements in the boreholes because sonic times obtained through the integration of sonic logs
usually differ from those obtained by means of a seismic pulse (well velocity surveys/ check shot
surveys/ VSP). The reasons for drift, range from basic discrepancies between two approaches-
geometry, frequency etc to more trivial disturbances to sonic readings caused by cycle skipping,
detection of mud arrivals in large holes, formation alteration and/or invasion ( Strike, 1971, Gretener,
1961 and Goetz, etal, 1979).

Sonic measurements are susceptible to errors due to a variety of factors:

 Detection triggered by noise before the signal arrives.
 Hole enlargement in which case mud signal will arrive first.
 Signal stretch resulting in increased travel time.
 Cycle skip or negative stretch, when the signal recorded by near receiver is weaker than the
signal recorded by far receiver.
 Difference in frequency between signals emitted by seismic source ( ~ 50 hz ) and the sonic
signals (around 20 to 25 KHz), speed of sound being dependent on frequency.
 Formations, which have lower sonic velocity than mud due to presence of gas or poor
compaction.
 Filtrate invasion which, when it replaces a lighter slower fluid (such as oil or specially gas,
results in an increase of velocity in invaded zone in relation to the real velocity in virgin zone.
 High apparent dips, causing acoustic signal to follow a refracted path, which is shorter than
one along the borehole.

In spite of these reasons, either sonic or check shot measurement may be wrongly made, due to
technological inadequacies or malfunctions. Since the check shot / VSP method is simpler and agrees
with reflection survey results, sonic is more susceptive. Therefore, it is necessary to calibrate the sonic
logs in order to eliminate as many of these possible errors to use it for seismic application in a
particular area. This must be done before converting the depth scale to time scale using time depth
relationship established through sonic logs. Calibration is done either by using check shots or by using
VSP technique.

Interval velocity, average formation velocity are computed from both one way seismic time and
integrated sonic time data available in sixteen wells spread all over Upper Assam oil fields. Time,
velocity and drift curves are presented separately for each well. Average and interval velocity curves
are presented in same scale with different colors for seismic and sonic derived values. Drift curves
derived from check shot data available in four wells are also presented. All data are also given in the
tabular form for reference.

2
 SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

METHODOLOGY

 Seismic time is the check shot or VSP time corrected to vertical and reduced to the seismic
reference datum. Seismic Reference Datum (SRD) for the area under study is 91 meters above
mean sea level. Sonic time i.e. integrated travel time (ITT) is available as a function of depth i.e.
some times referred as log depth with K. B. as reference datum. Seismic one-way / two way time
is generally available at SRD.

 Seismic time SRD depths are converted to log depth by adding (K.B – SRD) to SRD depths. One
way sonic time is obtained from ITT pips on sonic log for corresponding SRD depths converted as
log depths. Starting value of ITT is taken as value of seismic time at a tie point depth.

 Average and interval velocities at particular depth are computed from both sonic and seismic time
using following equations.

Vavg = depth / time and Vint = ( depth2 –depth1 ) / ( time2 – time1)

 Drift is defined as the difference between seismic time and integrated sonic time i.e. Drift =
Seismic time – Sonic time. At tie point both seismic time and integrated sonic time are made
equal. In present case, both are made equal to seismic time and then drift is computed at every
VSP depth converted to log depth as mentioned earlier. The successive values obtained are then
plotted as a function of log depth, which produces drift curve. For a given pair of depths, difference
in drift between the two depths is the correction, which must be applied to the sonic travel time.

 Now zones are chosen, in which the character of the sonic interval velocity is about constant or
according to stratigraphy viz. Tipams, Barail Coal Shale or Barail Main Sand, Kopili etc. In each
zone, drift points are fitted by a segment of straight line. These segments are joined at knees,
which form the common boundaries between one zone to the next.

 The drift curves are used to adjust the sonic logs. On the drift curves, the slope of the segment of
the straight line joining two consecutive knees is the gradient of the drift, expressed in
microseconds/ft is the average drift, i.e. correction to be applied to the sonic transit time.

 When slope is negative, drift is said to be negative, which means that sonic time is longer than
seismic time. Finally, a corrected sonic transit time curve is to be obtained by applying respective
drift value for a particular zone.

RESULTS

Interval velocity, average formation velocity are computed from both one way seismic time and
integrated sonic travel time data available in thirteen wells (Fig. No. 1) spread all over Upper Assam oil
fields. Velocity and drift curves are presented separately in the figures 2 – 13 for each well. Average
and interval velocity curves are presented in same scale with different colors for seismic ( pink and red
) and sonic (light blue and blue) derived values. All the depths presented in figures are log depths in
meters if it not mentioned otherwise.

A linear increase of average velocity with depth is noticed in all wells. Velocity value derived from one-
way time are generally higher than those derived from sonic log. One anomalous thing is noticed that
interval velocity values derived from seismic one way time are showing finer variations, which are
some times cyclic in nature i.e. higher velocity followed by lower one, as compared to interval velocity
values derived from sonic log. Such cyclic variations are not seen on other log measurements as well.

3
 SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

On the drift curves, the slope of the segment of the straight line joining two consecutive knees is the
gradient of the drift, expressed in microseconds/ft is clearly indicated for different formations on these
figures. Well wise cumulative drift values in milliseconds for different intervals and drift values in
microseconds/ft to be applied to correct sonic transit time values (DeltaT) against different formations
in these wells are summarized in following tables. Zone wise block shifting of sonic log as per this
table will ensure integrated travel time (ITT) at par with seismic one way time by preserving formation
signatures required for seismofacies mapping.

CUMULATIVE DRIFT VALUES IN MILLISECONDS FOR INTEGRATED TRANSIT TIME

Sl. No. Well No. Log Depth Interval Total drift Values
( Meters ) (Milliseconds)
1. Chakimukh # X 1700 - 4485 -175
2. Changmaigaon # X 1650 - 3150 - 40
3. Charali # XX 1725 - 3775 -26
4. Demalgaon # X 540 - 3440 -101
5. Demalgaon # XX 1925 - 3750 -15
6. Geleki # XXX 1475 - 4350 - 123
7. Kuargaon # X 2400 - 4000 - 45
8. Lakwa # XX 2375 - 3550 - 35
9. Laxmijan # X 500 - 4000 - 101
10. Nazira # XX 1425 - 3381 - 86
11. Panidihing # X 1550 - 3875 - 18
12. Safrai # X 1771 - 4346 - 172
13. Sonari # X 710 - 3510 - 94

Anomalous negative drift on ITT in Chakimukh # X ( -175 msec), Geleki # XXX ( - 123 msec. ) and
Safrai # X ( -172 msec.) is noticed and more than half of the drift is attained in Girujan Clays itself.
The same phenomenon is observed in all the wells. In Charali # XX, both sonic and seismic travel
time are almost matching.

DRIFT VALUES ( MICRO SEC / FT ) AGAINST DIFFERENT FORMATIONS FOR CALIBRATION OF SONIC TRANSIT
TIME ( DT VALUE)

Sl. Well No. Girujan Tipams Tipams BCS BMS Kopili

No. Clays ( 1 – 3) ( 4 – 6)
1. Chakimukh # X - 26.38 - 12.85 - 12.06 - 16.04
2. Changmaigaon # X - 22.06 - 3.42 - 7.30
3. Charali # X - 10.21 - 5.62 - 5.41 + 3.24 - 9.41
4. Demalgaon # X - 19.94 - 3.84 - 2.25
5. Demalgaon # XX - 4.18 - 3.68 - 1.57
6. Geleki # XXX - 30.42 - 6.50 - 2.24 - 18.65 - 2.13
7. Kuargaon # X - 8.71 - 10.69 - 6.90
8. Lakwa # XX - 6.53 - 10.85
9. Laxmijan # X - 5.80 - 15.21 - 11.78 - 16.52
10. Nazira # XX - 14.31 - 29.31 - 2.83 - 1.30
11. Panidihing # X - 12.29 - 8.01 - 8.01 - 6.31 - 2.15 - 3.5
12. Safrai # X - 36.81 - 15.47 - 18.01 - 8.20 - 4.17
13. Sonari # X - 12.57 - 12.37 - 5.10

4
 SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

DISCUSSION

Product of density and sonic velocity measurements obtained in a borehole is generally used to derive
acoustic impedance curve, which in turn is used to generate a reflectivity log or reflection coefficient
defined as relative contrast of acoustic impedances of two successive zones marked with boundary as
acoustic discontinuity. Convolution of reflectivity log with a source wavelet produces synthetic seismic
trace or seismogram. When this is compared with the real seismic trace, it becomes possible to depth
match the real trace and to interpret it in terms of seismofacies. This is how well logs contribute to the
interpretation of seismic profiles and thus provides an essential link between surface geophysics and
geology.

Data provided by the density and sonic tools can not be used in their raw form, as log measurements
are subject to the influence of parasitic factors such as an invaded zone or bad hole conditions, and
must first be corrected before calculating the acoustic impedance reflection coefficients. Density log
responds mainly to invaded zone due to its shallow depth of investigation. Density value will be too
high in case of replacement of light hydrocarbon especially gas, and it must be corrected to give
density of virgin zone. However, in case of Upper Assam oil fields, hydrocarbon effect is not seen on
density measurement and does not require correction for invasion of mud filtrate on it. As such
calibration of density log is not considered in this report.

Well log editing has two measure facets, first to recognize that incorrect data have been recorded and
second to determine better values to substitute for the incorrect (Ausburn, 1977). In present case,
sonic log against washed/caved shale are not corrected since no appropriate velocity depth relation is
presently available for shale in the study area. Another point of establishing this relationship is more
difficult task in absence of unwashed shale zones in the available logs.

Many authors discuss physical and technological limitations of sonic log measurements in the past
(Dupal, etal, 1977, Thomas, 1984, Peyret and Mons, 1984). Negative and positive drifts are explained
as cycle skipping and dispersion of velocity respectively. It is very easy to explain the negative drift
observed against most of the formations in Upper Assam oil fields. Anomalous negative drift on ITT in
Chakimukh # X ( -175 msec), Geleki # XXX ( - 123 msec ) and Safrai # X ( -172 msec) is noticed and
more than half of the drift is attained in Girujan Clays itself. The same phenomenon is observed in all
the wells. Due to the swelling nature of the clays and sonic log is reading anomalously higher value of
transit time. Sonic tool is not able to read virgin zone velocity due to its shallow depth investigation.
Positive drift is some times noticed in Barail Coal Shale may be attributed to the o invasion of mud
filtrate causing higher velocity in comparison to the virgin zone having gas/ volatile material in cleat
porosities.

Although sonic measurements with sample interval of 6” with a higher resolution up to 2 feet, interval
velocity values derived from seismic one way time are showing finer variations, which are some times
cyclic in nature i.e. higher velocity followed by lower one, as compared to interval velocity values
derived from sonic ITT. These variations will result in creating fictitious reflections on synthetic
seismograms, which are not seen on log measurements. This is probably due to the early or late
picking of arrival of first arrival value on VSP due to its flat nature / low frequency wavelet ( ~ 50 Hz)
as compared to sonic wavelet ( ~ 20 - 25 KHz). Spurious nature is further confirmed by the absolute
value of formation velocity up to 8.0 km/sec, which is not possible in nature ( Fig. No. 36, 42 and 45 ).

There has been a common practice to generate continuous velocity log derived from counting of ITT
pips on sonic log within a successive interval of 10 mts. Velocity values plotted as function of ITT in a
compatible scale of one way time seismic section, provides an alternate of synthetic seismogram for
seismofacies mapping within major sequences. As this method is based on the concept of displaying
of sonic velocity as function of ITT in seismic one way time scale, drift values i.e. corrections are
presented for both in the present report for the generation of CVL for seismofacies mapping.

5
 SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

CONCLUSIONS

 Negative drifts for the calibration of ITT as well as sonic transit time, are generally observed in
all the wells. Amount of drifts is more in Girujan Clays. Actually, more than half of the total drift
is attained in Girujan Clays itself.

 Cumulative drift values in milliseconds to be applied for the calibration of ITT and drift values
in microseconds/ft to be applied to correct sonic Delta T values against different formations in
different wells presented in tabular form to correct the sonic log for seismic applications.

 Although sonic measurements with sample interval of 6” with a higher resolution up to 2 feet,
interval velocity values derived from seismic one way time are showing finer variations, which
are some times cyclic in nature i.e. higher velocity followed by lower one, as compared to
interval velocity values derived from sonic integrated travel time. These variations will result in
creating fictitious reflections on synthetic seismograms, which are not seen on log
measurements.

 Block shifting of sonic log ensures integrated travel time at par with seismic one way time by
preserving formation signatures required for seismofacies mapping. It is worth mentioning that
sonic log has not lost even a bit of its grace as continuous velocity log even with recording of
well seismic data/VSP.

ACKNOWLEDGEMENTS

Authors are thankful to ONGC authorities for their kind permission to present and publish this article.
Technical discussions with Shri J. L. Srivastava, Dy. G.M.(W), Shri U. C. Bhatt, Dy.G.M.(W) and other
colleagues of Well Logging Services, ONGC, Nazira, Assam, are gratefully acknowledged. Critical
review of the report by Shri Alok K. Bhanja, is thankfully acknowledged. Special thanks are due to I/C
RCC, GSD, ONGC, Jorhat, for providing VSP data for the present study.

REFERENCES

Anstey, N. A., 1977, Seismic Interpretation, the physical aspects.

th
Ausburn, B. E., 1977, Well log editing in support of detailed seismic velocities, SPWLA 18 Annual
Symposium.
Dupal, L., Gartener, J., and Vivet, B., 1977, Seismic applications of well logs. Fifth European Logging
Symposium, Oct. 20-21, 1977.
Goetz, L. and Dupal, L., 1979, An investigation into discrepancies between sonic log seismic check
shot velocities, technical paper Schlumberger.
Gretener, P. E. F., 1961, An analysis of the observed time discrepancies between continuous and well
velocities surveys; Geophysics, vol. 26, p 1-11.
Peyret, O. and Mons, F., 1984, Sonic vs. seismic velocities positive drift studies recording frequency
effect; Schlumberger Technical Paper.
Stick, E., 1971; An explanation of observed time discrepancies between continuous and well velocity
surveys. Geophysics, vol-36, p.285-295.
Thomas, D. H. 1984, Seismic application of Sonic Logs, Technical paper British Petroleum Co. Ltd.
Yadav, L., Singh A. K., Raju, P.B. and Bansal C.R., 1985, Study of sonic velocity in Upper Assam
Valley, ERBC , ONGC ,Nazira , Unpublished Report.

6
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

7
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 2 Velocity plots and drift analysis of Chakimukh # X

8
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 3 Velocity plots and drift analysis of

Changmaigaon# X

9
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 4 Velocity plots and drift analysis of

Charali# XX

10
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 5 Velocity plots and drift analysis of

Demulgaon# X

11
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 6 Velocity plots and drift analysis of

Geleki#XXX

12
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 7 Velocity plots and drift analysis of

Kuargaon# X

13
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 8 Velocity plots and drift analysis of

Lakwa # XX

14
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 9 Velocity plots and drift analysis of

Laxmijan # X

15
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 10 Velocity plots and drift analysis of Nazira # XX

16
SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 11 Velocity plots and drift analysis of Panidihing # X

17
 SOCIETY OF PETROLEUM GEOPHYSICISTS, SPG-2004

Fig.No. 12 Velocity plots and drift analysis of Safrai # X

18

View publication stats