19.9 Other Geophysical Monitoring
Methods
Parts of this section are taken from “Future challenges
and unexplored methods for 4D seismics”, by Landrø,
Recorder (2005).
Up to now, 4D seismic has proved to be the most
effective way to monitor a producing reservoir. However, as discussed above, there are several severe
limitations associated with time lapse seismic. One
of them is that seismic reflection data is sensitive to
acoustic impedance (velocity times density). Although
4D time shift can reveal changes in average velocity
between two interfaces, an independent measurement
of density changes would be a useful complement to
conventional 4D seismic. The idea of actually measuring the mass change in a reservoir caused by hydrocarbon production has been around for quite some
time. The limiting factor for gravimetric reservoir
monitoring has been the repeatability (or accuracy)
of the gravimeters. However, Sasagawa et al. (2003)
demonstrated that improved accuracy can be achieved
by using 3 coupled gravimeters placed on the seabed.
This technical success led to a full field programme at
the Troll Field, North Sea. Another successful field
example was time-lapse gravity monitoring of an aquifer storage recovery project in Leyden, Colorado
(Davis et al. 2005), essentially mapping water influx.
Obviously, this technique is best suited for reservoirs
where significant mass changes are likely to occur,
such as water replacing gas. Shallow reservoirs are
better suited than deep. The size of the reservoir is a
crucial parameter, and a given minimum size is
required in order to obtain observable effects. In
quite another field, gravimetric measurements might
help in monitoring volcanic activity by distinguishing
between fluid movements and tectonic activity within
an active volcano.
Around 2000 the hydrocarbon industry started to
use electromagnetic methods for exploration. Statoil
performed a large-scale research project that showed
that it is possible to discriminate between
hydrocarbon-filled and water-filled reservoirs from
controlled
source
electromagnetic
(CSEM)
measurements. Since this breakthrough of CSEM
surveys some years ago (Ellingsrud et al. 2002), this
technique has mainly been used as an exploration tool,
in order to discriminate between hydrocarbon-filled
rocks and water-filled rocks. Field tests have shown
that such data are indeed repeatable, so there should
definitely be a potential for using repeated EM surveys
to monitor a producing reservoir. So far, frequencies
as low as 0.25 Hz (and even lower) are being used, and
then the spatial resolution will be limited. However, as
a complementary tool to conventional 4D seismic, 4D
EM might be very useful. In many 4D projects it is
hard to quantify the amount of saturation changes
taking place within the reservoir, and time-lapse EM
studies might be used to constrain such quantitative
estimates of the saturation changes. Furthermore,
unlike conventional 4D seismic which is both pressure
and saturation sensitive, the EM technique is not very
sensitive to pressure changes, so it may be a nice tool
for separating between saturation and pressure
changes.
In another paper, Johansen et al. (2005) show that
the EM response over the Troll West gas province is
significantly above the background noise level, see
Fig. 19.23. If we use the deviation from a smooth
response as a measure of the repeatability level, a
relative amplitude variation of approximately 0.1–0.2
(measured in normalised EM amplitude units) is
observed. Compared to the maximum signal observed
at the crest of the field (2.75) this corresponds to a
repeatability level of 4–7%, which is very good compared to conventional time-lapse seismic. This means
it is realistic to assume that time-lapse EM data can
provide very accurate low-resolution (in x-y plane)
constraints on the saturation changes observed on 4D
seismic data. Recently, Mittet et al. (2005) showed
that depth migration of low frequency EM data may
be used to enhance the vertical resolution. In a field
data example, the vertical resolution achieved by this
type of migration is maybe of the order of a few
hundred metres. The EM sensitivity is determined by
the reservoir thickness times the resistivity,
underlining the fact that both reservoir thickness and
reservoir resistivity are crucial parameters for 4D
CSEM. This means that the resolution issue with the
active EM method is steadily improving, and such
improvements will of course increase the value of
repeated CSEM data as a complement to conventional
4D seismic.
By using the so-called interferometric synthetic
aperture radar principle obtained from orbiting
satellites, an impressive accuracy can be attained in
504
M. Landrø
Methods
Parts of this section are taken from “Future challenges
and unexplored methods for 4D seismics”, by Landrø,
Recorder (2005).
Up to now, 4D seismic has proved to be the most
effective way to monitor a producing reservoir. However, as discussed above, there are several severe
limitations associated with time lapse seismic. One
of them is that seismic reflection data is sensitive to
acoustic impedance (velocity times density). Although
4D time shift can reveal changes in average velocity
between two interfaces, an independent measurement
of density changes would be a useful complement to
conventional 4D seismic. The idea of actually measuring the mass change in a reservoir caused by hydrocarbon production has been around for quite some
time. The limiting factor for gravimetric reservoir
monitoring has been the repeatability (or accuracy)
of the gravimeters. However, Sasagawa et al. (2003)
demonstrated that improved accuracy can be achieved
by using 3 coupled gravimeters placed on the seabed.
This technical success led to a full field programme at
the Troll Field, North Sea. Another successful field
example was time-lapse gravity monitoring of an aquifer storage recovery project in Leyden, Colorado
(Davis et al. 2005), essentially mapping water influx.
Obviously, this technique is best suited for reservoirs
where significant mass changes are likely to occur,
such as water replacing gas. Shallow reservoirs are
better suited than deep. The size of the reservoir is a
crucial parameter, and a given minimum size is
required in order to obtain observable effects. In
quite another field, gravimetric measurements might
help in monitoring volcanic activity by distinguishing
between fluid movements and tectonic activity within
an active volcano.
Around 2000 the hydrocarbon industry started to
use electromagnetic methods for exploration. Statoil
performed a large-scale research project that showed
that it is possible to discriminate between
hydrocarbon-filled and water-filled reservoirs from
controlled
source
electromagnetic
(CSEM)
measurements. Since this breakthrough of CSEM
surveys some years ago (Ellingsrud et al. 2002), this
technique has mainly been used as an exploration tool,
in order to discriminate between hydrocarbon-filled
rocks and water-filled rocks. Field tests have shown
that such data are indeed repeatable, so there should
definitely be a potential for using repeated EM surveys
to monitor a producing reservoir. So far, frequencies
as low as 0.25 Hz (and even lower) are being used, and
then the spatial resolution will be limited. However, as
a complementary tool to conventional 4D seismic, 4D
EM might be very useful. In many 4D projects it is
hard to quantify the amount of saturation changes
taking place within the reservoir, and time-lapse EM
studies might be used to constrain such quantitative
estimates of the saturation changes. Furthermore,
unlike conventional 4D seismic which is both pressure
and saturation sensitive, the EM technique is not very
sensitive to pressure changes, so it may be a nice tool
for separating between saturation and pressure
changes.
In another paper, Johansen et al. (2005) show that
the EM response over the Troll West gas province is
significantly above the background noise level, see
Fig. 19.23. If we use the deviation from a smooth
response as a measure of the repeatability level, a
relative amplitude variation of approximately 0.1–0.2
(measured in normalised EM amplitude units) is
observed. Compared to the maximum signal observed
at the crest of the field (2.75) this corresponds to a
repeatability level of 4–7%, which is very good compared to conventional time-lapse seismic. This means
it is realistic to assume that time-lapse EM data can
provide very accurate low-resolution (in x-y plane)
constraints on the saturation changes observed on 4D
seismic data. Recently, Mittet et al. (2005) showed
that depth migration of low frequency EM data may
be used to enhance the vertical resolution. In a field
data example, the vertical resolution achieved by this
type of migration is maybe of the order of a few
hundred metres. The EM sensitivity is determined by
the reservoir thickness times the resistivity,
underlining the fact that both reservoir thickness and
reservoir resistivity are crucial parameters for 4D
CSEM. This means that the resolution issue with the
active EM method is steadily improving, and such
improvements will of course increase the value of
repeated CSEM data as a complement to conventional
4D seismic.
By using the so-called interferometric synthetic
aperture radar principle obtained from orbiting
satellites, an impressive accuracy can be attained in
504
M. Landrø