Monitoring overburden changes on the Valhall Field

1
Monitoring overburden changeson the Valhall Field
T. Røste1, M. Landrø1, and P. Hatchell2 1NTNU,
Trondheim, Norway 2Shell EP, Rijswijk, The
Netherlands ROSE-meeting, April 2006
2
Objective

• Discriminate between changes in thickness and
  velocity in the overburden of the Valhall Field
• (from time-lapse OBC data, surveys 1 and 3)
• Show examples of observed distortion zones
  (Valhall) correlated with buried faults

3
Time-lapse time shifts capture changes in both
thickness and velocity
The challenge is to discriminate between ?z and
?v!
4
Methodology
5

• We search for the a-value leading to minimum
  least square error between estimated and picked
  relative time shifts

6

• When best fitted a-value is found,
• the changes in layer thickness and velocity are
  given by

7
Synthetic time-lapse example
Picked ?T/T
Based on a one layer model
8
Synthetic time-lapse example
Difficult to determine a-values below -5
Optimal
Estimated ?T/T vs offset for varies values of a
Based on a one layer model
9
Field data example Valhall
The selected LoFS 2D line (red) follows one of
the cables
10
LoFS Good quality and high repeatability
Position (km)
0
0.5
1.0
1.5
2.0
Time (s)
Top Res
Survey 1
Survey 3
11
Estimated ?T0/T0 for Top Res horizon (LoFS)
Red x Initial estimation Blue Smoothed
12
Several positions show a -5 for the sequence
SeaBed TopRes
Blue solid Best fitted a
Red x Picked relative time shifts, ?T/T.
Dashed Initial and last guess of a.
13
More examples indicating a -5 for the sequence
SeaBed TopRes
Blue solid Best fitted a
Red x Picked relative time shifts, ?T/T.
Dashed Initial and last guess of a.
14
Some positions show low sensitivity in a
Difficult to determine a
Red x Picked ?T/T. Dashed Initial and last
guess of a.
15
Standard deviation in a versus two key
parameters a itself and standard deviation in
picked time shifts
16
Results for all positions for sequence SeaBed
TopRes
Uncertainties given in dashed lines.
17
Time-lapse distortions (time shifts) moving with
offset
(From a different line)
18
Time shifts in offset vs position domain
19
Coherency Time-slice at 1960 ms
(Valhall streamer data)
20
Coherency Time-slice at 1960 ms
N
m
2 Faults Younger red fault (strike slip) cuts
the older yellow fault
(Valhall streamer data)
21
RMS Amplitude Changes at TopRes (LoFS 1-3)
Strong indications of a growing fault at the SW
tip of the red fault
22
Inline 1441 showing red fault
23
Discussions and Conclusions

• LoFS data offers high degree of repeatability
• We estimate a -5 for the Valhall overburden
• Similar a-values are obtained by Hatchell et al.
  (2005)
• and Carcione et al. (2006)
• Subsidence of Top Reservoir horizon 0.5 m
  (max),
• Corresponding velocity decrease (from sea bed to
  Top Reservoir) 2.0 m/s
• Time-lapse distortion zones are observed and
  correlated to growing faults
• Such distortions cannot be detected from
  poststack analysis

24
Acknowledgments

• Total for financial support
• Valhall partnership (BP-op, Shell, Amerada Hess,
  and Total) for providing and permission to use
  and present the LoFS data
• Peter Wills for valuable help and discussions
• This work was done at Shell EP (Rijswijk).
• Thanks to the 4D group in Shell for hospitality
  and assistance in data analysis.
• The presented results and opinions do not
  necessarily reflect the view of the Valhall
  partnership