Download a new understanding of fluid-rock deformation

EUROCK 2006 – Multiphysics Coupling and Long Term Behaviour in Rock Mechanics –
Van Cotthem, Charlier, Thimus & Tshibangu (eds)
© 2006 Taylor & Francis Group, London, ISBN 0 415 41001 0
The New Geophysics: a new understanding of fluid-rock deformation
Stuart Crampin
School of GeoSciences, The University of Edinburgh, Scotland, UK
ABSTRACT: Recent observations of stress-aligned shear-wave splitting (seismic-birefringence) show that the
splitting is controlled by the stress-aligned fluid-saturated inter-granular microcracks and preferentially orientated
pores pervasive in most in situ igneous, metamorphic, and sedimentary rocks in the earth’s crust. These fluidsaturated microcracks are the most compliant elements of the rock-mass and control rock deformation. The
splitting shows that microcracks are so closely spaced that they verge on fracture-criticality and fracturing, and
are critical systems with the butterfly wing’s sensitivity. As a result, the evolution of such fluid-saturated rock
can be modelled with anisotropic poro-elasticity (APE), so that behaviour can be monitored with shear-wave
splitting, future behaviour calculated (predicted ), and in principle future behaviour controlled by feed back.
This is likely to have massive implications for virtually all solid-earth deformation at levels of stress below those
at which fracturing takes place.
1 INTRODUCTION
Widespread observations of stress-aligned shearwave splitting (seismic birefringence) in hydro-carbon
reservoirs, and above small earthquakes, show that
almost all in situ rocks are pervaded by distributions of
stress-aligned fluid-saturated microcracks (Crampin,
1994, 1999; Alford, 1986). These are grain-boundary
cracks in crystalline rocks, and preferentially-aligned
pores and pore throats in sedimentary rocks (Figure 1).
The degree (%) of observed splitting indicates that
microcracks are so closely spaced they verge on fracture criticality, when levels of cracking are so great
that shear-strength is lost and fracturing necessarily
occurs (Crampin, 1994, 1999). Phenomena verging
on criticality are critical systems are said to have
self-organised criticality (Jensen, 1998).
Critical systems are a New Physics (Bruce &
Wallace, 1998), a New Geophysics, with different
properties from conventional sub-critical behaviour.
Below criticality, individual elements of the system,
in our case the fluid-saturated microcracks interact
locally with conventional behaviour. At criticality, the
elements throughout the region combine and act cooperatively, and the rocks (sometimes over very large
volumes) spontaneously fracture leading to deterministic chaos. The behaviour can be calculated but the
final position depends critically on minuscule differences in initial conditions. Such behaviour at criticality
is characteristic of almost all complex complicated
systems which are extremely common. It is said to be
“one of the universal miracles of nature” that at critical
points whole assemblages of elements “are capable
of organising themselves into patterns of cooperative
activity” (Davis, 1989). Thus our claim that the cracks
in the crust are critical systems merely confirms that
the Earth behaves like almost every other complex
interactive system: the weather; quantum mechanics;
super-fluidity and super-conductivity; life cycle of
fruit flies; clustering of traffic; etc.
This brief note outlines the evidence and implications of the New Geophysics.The detailed deformation
of the crust can be monitored with shear-wave splitting; evolution calculated if the stress changes are
known; future behaviour predicted; and, in some circumstances, future behaviour potentially controlled by
feedback.
2
SHEAR-WAVE SPLITTING
The key phenomenon for observing the internal microcracked structure of rocks is stress-aligned shear-wave
splitting, where shear-waves split into two approximately orthogonal polarisations where the faster polarisation is aligned approximately in the direction of the
maximum horizontal stress (Figure 1) (Crampin, 1994,
1999). Seismic P-waves are rather insensitive to fluidsaturated microcracks, whereas stress-aligned shearwave splitting is a second-order effect that, by rotating
seismograms into preferential orientations, can typically be measured with first-order accuracy. Thus
measurements of shear-wave splitting often have much
greater resolution than conventional seismics and it
is possible to show that the parameters that control
shear-wave splitting are directly those that control
539
Figure 1. Schematic illustration of stress-aligned seismic
shear-wave splitting in the stress-aligned microcracked crust.
Figure 3. The Anisotropic Poro-Elastic (APE) model of
rock deformation where the driving mechanism is driven
by pore-fluid movement along pressure-gradients between
neighboring cracks at different orientations to the stress field.
Figure 2. Schematic illustration of observed percentage of
shear-wave velocity anisotropy in terms of uniform distributions of penny-shaped cracks.
low-level deformation before fracturing occurs
(Crampin, 1999).
Crack density ε is a dimensionless parameter
approximately equal to Na3 /v, where N is the number of cracks of radius a in volume v (Crampin, 1994).
As crack density ε is approximately equal to one hundredth of the percentage of the maximum shear-wave
velocity anisotropy, it is comparatively easy to estimate crack density from observations of shear-wave
splitting. The crack density determines the effect of
cracks on the effective elastic tensor of an anisotropic
microcracked solid.
The observed shear-wave velocity anisotropy, 1.5%
to 4.5%, illustrated schematically in Figure 2, indicates
a narrow range 0.015 ≤ ε ≤ 0.045 of the inferred crack
density in ostensibly intact rock. The values appear
to be more-or-less independent of porosity, geology,
and tectonic history, and are observed to have similar
values in all sedimentary, igneous, and metamorphic
rocks, with only a few well-understood exceptions
(Crampin, 1994, 1999). Since there is a factor of less
than two in average crack radius, between the minimum observed, and the maximum at fracture criticality, almost all in situ rocks are so heavily microcracked
that shear strength lost and fracturing occurs in the
presence of almost any increase in differential stress.
The limited range of crack density means that
distributions of cracks in in situ rocks are critical
systems verging on criticality and failure and have
self-organised criticality (SOC) (Jensen, 1998).
The evolution of such fluid-saturated microcracks
can be modelled by Anisotropic Poro-Elasticity (APE),
where the mechanism for deformation is fluid movement by flow or dispersion along pressure gradients
between neighbouring grain-boundary cracks in crystalline rocks, and aligned pores and pore-throats in
sedimentary rocks, at different orientations to the
stress field (Zatsepin & Crampin, 1997; Crampin &
Zatsepin, 1997). APE is a fully three-dimensional
model but 3D variations are difficult to plot on paper,
so Figure 3 is a schematic image of the APE mechanism on an initially random distribution of vertical
cracks for increasing differential stress sH .
3 APE MODEL OF ROCK MASS EVOLUTION
APE modelling approximately matches a huge range
of different phenomena referring to cracks, stress, and
shear-wave splitting in both exploration and earthquake seismology (Crampin, 1999, 2004; Crampin &
Chastin, 2003). The match is approximate because
the internal microcrack structure of in situ rocks is
essential inaccessible. Even drill cores deform rock:
drilling cools and de-stresses in situ rock, and disturbs
the compliant internal crack geometry by invasion of
high-pressure drilling fluids.
The most successful in situ calibration to-date
(Figure 4) is Angerer et al. (2002) who modelled
changes in three-component reflection surveys in a
hydrocarbon reservoir following both high-pressure
and low-pressure CO2 -injections in a fractured carbonate reservoir at 1280 m-depth.
The first five traces in Figure 4(a) show observed
three-component record sections rotated into faster
540
These are the best in situ demonstrations of the applicability of APE-modelling. This means that, if the
changes to the rock are known, the effects can, at
least in some circumstances, be calculated (predicted)
by APE. It also means that if a particular result is
desired: a particular degree of aligned cracking during
a water-flood operation, say, the intended effect could
be controlled by adjusting fluid-pressure injections by
feedback.
4
Figure 4. (a) Pre-injection waveforms of a multi-component
nearly-vertical ray reflection survey near the centre of the
Vacuum Field, New Mexico, carbonate reservoir (Angerer
et al., 2002). S1, S2, and P are record sections with mutually orthogonal polarisations, where the horizontals S1, and
S2, have been rotated into the split shear-wave polarisations parallel and perpendicular to the direction of maximum
horizontal stress, respectively. Left-hand (LH) five traces
are observed waveforms at recorders 17 m apart, and the
right-hand (RH) three traces are synthetic seismograms modelled by APE to match the shear-wave arrivals. Top and
bottom of injection zone for shear waves are marked by
arrows with time-delays in ms/km. (b) Post-injection waveforms after a high-pressure CO2 -injection. Again, the LH
traces are observations and RH traces are synthetic seismograms modelled by APE with structure from (a) and an
injection pressure of 2500psi.
S1-wave and slower S2-wave horizontal, and vertical P-wave, components showing ∼2% shear-wave
velocity anisotropy. The last three traces show theoretical reflections (calculated by ANISEIS, Taylor, 2000)
through a model structure incorporating fractures and
microcracks. The shear-wave splitting arrivals from
the top and bottom of the target zone (arrowed)
match observations. Figure 4(b) shows the observed
arrivals after a high-pressure CO2 -injection, where
the modelled arrivals are the initial cracked model in
Figure 4(a) modified by inserting the fluid-injection
pressure into APE.
The match of observations and APE-model is excellent, as it is for a low-pressure injection (not shown).
90◦ -FLIPS OF SHEAR-WAVE
POLARISATIONS IN HIGH-PRESSURE
CRACK FLUIDS
Figure 4 also shows another confirmation of APEmodelling. In Figure 4(a) the two-way reflection times
of the (faster) S1-wave from the top and bottom of
the target zone differ by 176ms whereas, as expected,
the delay for the (slower) S2-wave is 178 ms. In contrast, in Figure 4(b) after the high-pressure injection,
with the same seismogram rotation, the S1-wave delay
is 204 ms and the S2-wave delay is 184 ms. Thus the
S2-wave is now faster than S1, and there has been a
90◦ -flip in shear-wave polarisations in the presence
of critically high pore-fluid pressures (Crampin et al.,
2004).These 90◦ -flips, in some circumstances can lead
to observations of time-delays between split shearwaves decreasing with distance, and observations of
shear-wave polarisations parallel to the minimum horizontal stress instead of the more typically parallel to
the maximum horizontal stress.
90◦ -flip phenomena, first modelled theoretically by
APE (Crampin & Zatsepin, 1997), have now been
observed elsewhere in critically-high pressured hydrocarbon reservoirs, and on all seismically-active fault
zones due to the high fluid-pressures necessary to
permit movement on lithostatically compressed faults
(Crampin et al., 2004). Above large faults, 90◦ -flips
are observed at the surface, but on smaller seismicallyactive faults, the 90◦ -flips lead to negative, say, timedelays near the fault, and the positive time-delays along
the normally-pressurised remainder of the ray path
to the surface. The summation of negative and positive time-delays leads to the large ±80% scatter in
observed shear-wave splitting time-delays above all
small earthquakes (Crampin et al., 2004).
5
INDIRECT CONFIRMATION OF
APE-MODELLING
Further indirect confirmation of APE-modelling is
that APE indicates that the principal effect of changes
of low-level stress is to modify microcrack aspect
ratios. In particular, increases of stress before fracturing occurs (pre-fracturing deformation) can be
541
monitored by increases in shear-wave splitting timedelays in particular segments of the solid angle of
ray path directions (Crampin, 1999). Observations of
such, rather subtle, easy-to-overlook changes, monitor the accumulation of stress before earthquakes
and allow estimates of the progress towards fracturecriticality and earthquake occurrence. The duration of
the increase allows the magnitude of the impending
earthquake to be estimated. Recognised with hindsight
before some 17 earthquakes (Volti & Crampin, 2003;
Gao & Crampin, 2004). On one occasion the time,
magnitude, fault break of an M 5 earthquake in Iceland was successfully stress-forecast three days before
the event (Crampin, 1999).
6 CRITICAL SYSTEMS OF
FLUID-SATURATED CRACKS
IN THE EARTH
between adjacent fluid-saturated grain-boundary
cracks and pore-throats. This is a quantifiable physical
process that can be modelled, monitored, and calculated by APE (Zatsepin & Crampin, 1997; Crampin &
Zatsepin, 1997; Crampin & Chastin, 2003).
The criticality, where different critical systems have
similar statistical behaviour near criticality, despite
very different sub-critical physics, is the reason
APE-modelling matches a huge range of phenomena
(Crampin & Chastin, 2003). This is known as criticalpoint universality (Bruce & Wallis, 1989), and implies
the self-similar scaling seen in crack distributions,
as well as the well-known Gutenberg-Richter earthquake magnitude frequency relationship, and many
other phenomena.
7
The capability of imaging the behaviour before earthquakes ranging from a M 1.7 in Iceland to the Ms
7.7 Chi-Chi earthquake by the highly-constrained APE
model and calculate, even predict, the response of
fluid-saturated microcracked rock to changing conditions in a highly complicated heterogeneous crust,
is remarkable and requires explanation (Crampin &
Chastin, 2003).
Critical systems are typically sensitive (the "butterfly wing’s" sensitivity) to otherwise negligible
variations in initial conditions that can lead to
order-of-magnitude differences as the systems evolve
(Crampin, 1999; Crampin & Chastin, 2003). This
manipulation of fluid-saturated microcracks as monitored by shear-wave splitting indicates that stressaligned fluid-saturated microcracks are remarkably
pervasive features with similar parameters in almost
all rocks in the crust. The observed calculability is
because the fluid-saturated cracks in the crust are so
closely spaced that they are critical systems (Bruce &
Wallis, 1989). Critical systems involve dynamic interactive processes that below criticality perturb only
locally. Once systems reach criticality, all members
of the critical system influence all other members
(Jensen, 1998; Bruce & Wallis, 1989; Crampin &
Chastin, 2003).
The transition temperature of equilibrium thermodynamics is the classic critical system, but critical systems are common in an enormous range of phenomena
including almost all complex interactive systems in
nature and elsewhere. Crampin & Chastin (2003) suggest that the stress-aligned fluid-saturated cracks in the
earth’s crust are also interactive critical systems. Similar schemes for the Earth have been suggested previously. The difference here is that we have identified the
micro-scale mechanism for deformation as stressinduced fluid-movement along pressure-gradients
OBSERVED SENSITIVITY
The expected sensitivity was confirmed by an experiment to test a borehole Stress-Monitoring Site (SMS)
(Crampin et al, 2003) for stress-forecasting earthquakes. The “SMSITES” SMS near the Húsavík-Flatey
transform fault of the Mid-Atlantic Ridge in Iceland recorded seismic travel times over 315 m at
500 m-depth (Figure 5). Repeated pulsing of the
Downhole Orbital Vibrator (DOV) borehole source
several times a minute over 13 days allowed, with
100-fold stacking, ±0.02 ms time-lapse accuracy in Pwave and split shear-wave (SV and SH ) travel times,
and shear-wave anisotropy (SV – SH ). The arrivals
showed well-recorded 5%, 2%, 2%, and 10% variations, respectively, which correlated with NS and
EW Global Position System (GPS) variations and
a one metre decrease in the level of a water well
immediately above the Húsavík-Flatey fault. All seven
variations coincided with small-scale seismicity, with
total energy release equivalent to one M 3.5 event, on
a neighbouring transform fault 70 km-away (Crampin
et al., 2003). These well-observed variations, show
exceptional sensitivity, well beyond that expected in
a conventional brittle-elastic crust at several hundred
times the conventional source diameter.
8
IMPLICATIONS OF THE NEW GEO-PHYSICS
FOR HYDROCARBON PRODUCTION
Since liquid-saturated cracks have little effects on
P-wave propagation, the New Geophysics is likely to
have little effect on conventional exploration seismology, where the principal tool is P-wave arrival times.
Consequently, if you are happy with conventional
P-wave seismics, and happy with conventional resolution, you can probably ignore this New Geophysics.
542
However, if you wish for higher resolution, and particularly wish to use shear-wave seismics for a better understanding of low-level deformation, you are
necessarily committed to the New Geophysics.
There are two major effects. The New Geophysics
potentially alters the behaviour of seismic wave propagation, and alters the geophysical behaviour of the
reservoir.
8.1 Monitoring production with time-lapse
seismics in Single-Well Imaging (SWI)
The movement of oil/water fronts can be analysed by
subtracting record sections before and after some oil
field production process in time-lapse seismics. Success depends on successive records shot with identical
source-to-receiver geometry. Such detailed sections
are expensive offshore, where many major oil fields
are situated, and one of the major expenses is laying areal arrays of sea-floor geophones for optimal
time-lapse seismics.
Figure 5 shows time-lapse signals contaminated
by substantial temporal variations in seismic travel
times over 315 m correlating with distant small-scale
seismicity. The effects of stress accumulation before
higher seismicity, or volcanic eruptions, are likely to
be seen across the whole tectonic plate (Crampin et al.,
2003). Note that the effects of the New Geophysics are
non-linear, as in the ±80% scatter of shear-wave timedelays observed above all small earthquakes (Crampin
et al., 2004), that cannot be interpreted by conventional
sub-critical techniques. Consequently, any in situ measurements necessarily degrade both temporally and
spatially as soon they are made, and the longer the
ray paths the greater the possible degradation.
The only way to avoid such degradation is to make
and interpret measurements at the time and place (the
producing reservoir) they are required by Single-Well
Imaging (SWI) (Crampin, 2004). SWI is where the
scattered reflections from a borehole (DOV) source
are recorded by three-component recorders behind
the casing or behind tubulars in the same well as
the source. There are time-lapse techniques for interpreting the scattered signals in terms of changing
background: moving oil/water contacts for example
(Crampin, 2004).
One great advantage of SWI is that it is much
cheaper (× 1/100) than conventional 4D reflection surveys. Since SWI tracking of fluid fronts is largely independent of geological structure, SWI could eventually
replace costly reflection surveys.
8.2 Slower production for More Oil REcovery
(SMORE)
Figure 5. Observations at the prototype SMS (Crampin
et al., 2003). (a), (b), and (c) are travel times in ms of P-,
SV - and SH -waves, and SV -SH anisotropy, respectively, at
500 m-depth between boreholes 315 m-apart. The directions
are parallel to and about 100 m South of the Húsavík-Flatey
Fault (HFF). Shear-waves are polarised SV, and SH propagating in a vertical symmetry direction of the stress field. Also
shown are (d) NS and EW (GPS) displacements in mm; and
(e) water-pressure ‘pulse’ measured in bars in a well immediately above the HFF (the ‘pulse’ is ∼1 m deep; the ∼40 cm
oscillations are oceanic tides). Observations (a to e), correlate with (f ), a histogram of small scale seismicity on a
parallel fault 70 km NNW of the SMS. The total energy of
the seismicity is equivalent to one magnitude M 3.5 event.
One of the characteristics of critical systems with SOC
is that the self-similarity and calculability only occur
when the complex interactions are responding to slow
changes (Jensen, 1998). Rapid changes and aggressive production strategies are likely to produce chaotic
deformation which will not be self- similar or calculable. This suggests that modelling and calculating the
response will only be possible for slow recovery rates,
hence the hypothesis of Slower production for MOre
REcovery (SMORE).
The question is how slow is slow and how much
more is more? The delay between the initial and second
survey in for Figure 4 was approximately two weeks
and was clearly sufficient as the results were calculable, but it could have been days or hours. Note that
the slow recovery might need to be for the whole production history. Initial aggression might so disturb the
rock as to be irreversible.
There are to our knowledge no estimates of how
much more is more. Currently, oil fields typically
recovery less than ∼40% of the oil in the reserve.
Thus, even a conservative increase of 2%, say, to 42%
(it might be substantially greater) of produced oil over
possibly marginally longer recovery would be ∼5%
more oil. This would be additional profit on the initial
543
infrastructure cost of oil rigs and pipe lines at the cost
of a slower recovery rate.
8.3 Optimising water-flooding by APE-modelling
The response of the rock mass can be potentially controlled by matching time-lapse shear-wave splitting in
4D reflection surveys or VSPs to APE-modelling by
feedback.
This might be important for fluid-injection. If a
water-flood operation were initially monitored by
appropriate seismic measurements, such as VSPs or
SWI, it would be possible to recognise whether cracks
had opened in the desired direction to aid fluid sweep
operation. If non-optimal orientations were indicated,
different injection procedures could be attempted and
tested.
This could establish the likelihood of success of the
water fluid operation within hours or days of the flood
and long before production returns indicated success
or failure.
9
CONCLUSIONS
The New Geophysics is a recent construct and the full
implications will take some time to evolve. Why do we
need a New Geophysics? After some 80+ years of oil
production and vast expenditure of research funds, we
still recover, typically, less than ∼40% of the total oil
in the reservoir. Why should we be so unsuccessful?
Shear-wave splitting (seismic birefringence) shows
there are stress-aligned fluid-saturated microcracks in
almost all in situ rocks that are so closely spaced
they are highly compliant and are critical systems.
This means that the reservoir and seismic waves do
not behave as expected in conventional geophysics.
Figures 4 and particularly Figure 5, for example,
show that rocks, including hydrocarbon reservoirs,
display sensitivity to minor disturbances, with much
more sensitivity than would be expected in a conventional brittle-elastic crust. These and other fundamental anomalies suggest that there is a serious gap in our
understanding of how rocks behave. The New Geophysics may not be the whole answer to our lack of
success in oil production, but it certainly provides a
large part of the answers, where the key observable is
stress-aligned shear-wave splitting.
The New Geophysics provides the opportunity
to: monitor with shear-wave splitting; model future
behaviour with anisotropic poro-elasticity (APE), predict future behaviour if changing conditions can be
estimated, and in appropriate circumstances control
future behaviour by feedback.
REFERENCES
Alford, R.M. 1986. Shear data in the presence of azimuthal
anisotropy: Dilley, Texas. 56th Ann. Int. SEG Meet.,
Houston, 1986, Expanded Abstracts 476–479.
∗
Angerer, E., Crampin, S., Li, X.-Y. & Davis, T.L. 2002.
Processing, modelling, and predicting time-lapse effects
of overpressured fluid-injection in a fractured reservoir.
Geophys. J. Int. 149: 267–280.
Bruce, A. & Wallace, D. 1989. Critical point phenomena:
universal physics at large length scale. In P. Davis (ed),
The New Physics. Cambridge: Cambridge Univ. Press:
236–267.
∗
Crampin, S. 1994. The fracture criticality of crustal rocks.
Geophys. J. Int. 118: 428–438.
∗
Crampin, S. 1999. Calculable fluid-rock interactions.
J. Geol. Soc. 156: 501–514.
∗
Crampin, S. 2004. The New Geophysics: implications
for hydrocarbon recovery and possible contamination of
time-lapse seismics. First Break 22: 73–82.
∗
Crampin, S. & Chastin, S. 2003. A review of shear-wave
splitting in the crack-critical crust. Geophys. J. Int. 155:
221–240.
Crampin, S. & Zatsepin, S.V. 1997. Modelling the compliance of crustal rock: II – response to temporal changes
before earthquakes. Geophys. J. Int. 129: 495–506.
∗
Crampin, S., Volti, T. & Stefánsson, R. 1999. A successfully
stress-forecast earthquake. Geophys. J. Int. 138: F1–F5.
∗
Crampin, S., Chastin, S. & Gao, Y. 2003. Shear-wave
splitting in a critical crust: III - preliminary report of
multi-variable measurements in active tectonics. J. Appl.
Geophys. 54: 265–277.
∗
Crampin, S., Peacock, S., Gao, Y. & Chastin. S. 2004. The
scatter of time-delays in shear-wave splitting above small
earthquakes. Geophys. J. Int. 156: 39–44.
Davis, P. 1989. The New Physics: a synthesis. In P. Davis (ed),
The New Physics. Cambridge: Camb. Univ. Press.
∗
Gao, Y. & Crampin, S. 2004. Observations of stress relaxation before earthquakes. Geophys. J. Int. 157: 578–582.
Jensen, H.J. 1998. Self-organized criticality. Cambridge:
Cambridge Univ. Press.
Taylor, D.B. 2000. ANISEIS Manual: Version 5.5. Edinburgh:
Macroc Ltd., 10 Jordan Lane, EH10 4RB.
∗
Volti, T. & Crampin, S. 2003. A four-year study of shearwave splitting in Iceland: 2. Temporal changes before
earthquakes and volcanic eruptions. In D.A. Nieuwland (ed), New insights into structural interpretation
and modelling. Geol. Soc. Lond., Spec. Publ. 212:
135–149.
Zatsepin, S.V. & Crampin, S. 1997. Modelling the compliance of crustal rock: I - response of shear-wave splitting
to differential stress. Geophys. J. Int. 129: 477–494.
∗
Available at http://geos.ed.ac.uk/homes/scrampin/opinion/.
544