SPE 103870

Improved Oil Recovery by Seismic Vibration: A Preliminary Assessment of Possible

Mechanisms
C. Huh, SPE, U. of Texas at Austin

Copyright 2006, Society of Petroleum Engineers

Another important question that needs to be addressed is
This paper was prepared for presentation at the First International Oil Conference and Exhibition how to deliver vibrational energy effectively to a large volume
in Mexico held in Cancun, Mexico, 31 August–2 September 2006.
of reservoir where the oil is. Fluid pressure oscillation damps
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
down quickly in porous media and even the rock deformation
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to wave attenuates significantly. One possible way is to transmit
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at rock displacement wave continually so that its accumulation
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
can form resonance in the reservoir. The resonance generation
for commercial purposes without the written consent of the Society of Petroleum Engineers is with the reservoir length scale as the resonance wavelength
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous requires a very slow wave, and recent literature on the subject
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
is discussed.

ABSTRACT INTRODUCTION
Use of seismic vibration has been proposed as a low-cost Increase in oil production from mechanical vibration has
IOR method, with reports of some successful field results. A been initially observed in Russian oil fields when earthquake
key difficulty with the method is that its mechanism for oil occurred nearby (Beresnev and Johnson 1991; Nikolaevskiy et
recovery is not yet known, and accordingly, process al. 1996). While some of the observations of production
performance prediction and reliable project design is so far not change were obviously due to the large-scale upheaval in
possible. In this study, a critical review of literature on the reservoir structure from the earthquake’s high energy input,
field and laboratory results and theoretical modeling work is the production changes have also been observed when the
first made. Employing a simple pore-level model and the distance between the epicenter and the reservoir was quite far,
well-established capillary number correlation, mobilization of so that the reservoir has been subjected to only low-frequency,
waterflood residual oil by seismic vibration is shown to be low-amplitude vibrations. Prompted by these natural
unlikely, confirming earlier coreflood data. occurrences, vibration has been deliberately applied to affect a
A more likely mechanism for IOR is that vibration reservoir’s oil production response. All manners of vibration
squeezes out the oil bypassed by earlier recovery methods, due were applied: (1) periodic “thumping” of the surface with a
to reservoir heterogeneity. While it is quite difficult to bring large weight to generate seismic waves (vibroseis); (2)
the injection fluids to the low-permeability zones in a highly injection of water in pulses from well to generate pressure
heterogeneous reservoir, vibrational wave can transmit waves in a water-hammer fashion; (3) sonic or ultrasonic
effectively through both high- and low-permeability zones. generation in the fluid phase at the wellbore; and (4) various
When a seismic vibration propagates through a heterogeneous modes of chemical/nuclear explosions either downhole or at or
reservoir, the pore pressure response in different permeability near the ground surface.
layers becomes different, generating a transient pressure In this paper, we will limit our consideration to the use of
difference and crossflow between the layers, potentially “seismic” vibrations which have low frequency (~1 to 100 Hz)
squeezing out oil from the low-permeability zones. The rock and low amplitude (~1 to 100 μm). This excludes the use of
deformation and fluid flow equations are coupled to describe sonic or ultrasonic waves (Roberts et al. 2000), which have
the effects of low-frequency oscillatory poroelastic frequencies considerably higher than the above, and cannot be
deformation on crossflow. Improved oil recovery due to transmitted for a long distance; and also excludes the chemical
vibration-induced crossflow is approximately estimated. Our or nuclear explosions which will generate very large-
assessment that seismic vibration recovers only the oil amplitude vibrations, probably causing irreversible geological
bypassed due to reservoir heterogeneity may explain why the structural changes. Also, we will only cite relevant references
method’s success is rather erratic, since accurate advance for the related process of the “pulsed” or “cyclic”
characterization of reservoir heterogeneity and the bypassed waterflooding, in which the water injection is carried out with
oil is inherently difficult. On positive side, seismic vibration periodic changes of rate (Li et al. 1995; Yuan et al. 1995;
may improve access of IOR chemical to low-permeability Surguchev et al. 2002; Stirpe et al. 2004; Booth and Araujo
zones such as matrix portion of naturally fractured reservoirs, 2005; Ivanov and Araujo 2006). Because the cycle period is
for better oil extraction from hard-to-access tight matrices. usually in days or months, its improved oil recovery
2 SPE 103870

mechanism, if it is indeed effective, would be different from delivered at the oil reservoir would be only a very small
the mechanism for the seismic vibration-based IOR. fraction of the energy generated at the vibration source. With
This paper has three objectives: (i) to review the available a continuous supply of vibration from the source, if a
literature on application of mechanical vibration for improved resonance generation is possible at the oil location, the small
oil recovery; (ii) to make a preliminary evaluation of various amounts of energy will accumulate to generate a large-
possible mechanisms for oil recovery by seismic vibration, amplitude wave. Recent theoretical studies suggesting such a
and (iii) to propose a potential mechanism for effective possibility are reviewed.
transmission of vibrational energy in subsurface formations.
The first objective is important because, despite the numerous LITERATURE REVIEW ON IOR BY SEISMIC
reports of field success, the effectiveness of the suite of VIBRATION
techniques for oil recovery has not been demonstrated with Subsequent to the earlier reviews cited above, other
detailed, supporting data. In particular, various oil recovery reviews on the vibration-based IOR have been published by
mechanisms attributed for the reported successes are generally Kouznetsove et al. (1998), Westermark et al. (2001),
not supported by laboratory observations or by appropriate Kuznetsov et al. (2002), Dobronravov (2002), and Roberts et
theoretical modeling studies. In view of the above al. (2003). While the above reviews describe the past field
uncertainty regarding the process mechanisms, our second activities on the vibration-based IOR, their description of the
objective is aimed at proposing a most likely oil recovery oil recovery mechanisms is quite insufficient, and filling this
mechanism, and suggests the conditions under which the gap is our main objective. In this section, therefore, we will
vibration application is likely to be effective in improving oil focus on review of laboratory experiments and modeling work
recovery. The third objective is important because it is reported in literature.
extremely difficult to deliver a large magnitude of vibrational Extensive field applications of vibration-based IOR have
energy to a large reservoir volume deep down the ground been made in Russia, as reported in the reviews cited above,
surface, because of attenuation. Without the delivery of the but the details of the techniques used and their outcome are
sufficient vibrational energy to where the oil exists, oil unfortunately not available in Western literature. In Appendix
displacement would not be possible, regardless of the local oil I, the results from a number of field cases (that are reported in
recovery mechanism. the English translation of the Russian and Chinese journals)
In the next Section, the literature on improved oil recovery are summarized. Recent field activities in U.S. and Canada
by seismic vibration will be reviewed in some detail, with are also briefly reviewed.
focus on laboratory experimental results and on the oil Even though it is difficult to draw definitive conclusions
recovery mechanisms proposed. This is because, as from the field results, some general observations can be made:
mentioned above, the field results are difficult to evaluate with (1) when water/oil ratio (WOR) is large, oil production
only limited data reported. Some representative field results increased; but when WOR is small, oil production increase is
reported in literature are nevertheless briefly reviewed in not as noticeable; (2) not all producers respond to vibration;
Appendix I. In the Section, “Preliminary Evaluation of Oil and (3) there appears to be a reservoir-specific frequency
Recovery Mechanisms”, it is first shown why the mobilization which is most effective for improved oil recovery. The first
of waterflood residual oil is not likely, by seismic vibration and second observations in the above suggest that the
that is realizable in the reservoir. It is argued, based on the increased oil production is probably from improved sweep of
well-established capillary number correlation and a simplified the oil bypassed due to reservoir heterogeneity, rather than
pore-level model, that the capillary force holding oil ganglia in from mobilization of the waterflood residual oil. The second
pores is much too strong for a realistic vibration amplitude to observation indicates that understanding the mechanisms for
overcome. It is then shown that, for a highly heterogeneous oil mobilization and recovery is quite important. The third
reservoir, the oil bypassed by earlier recovery methods and observation suggests the possibility of resonance generation at
still left in its low-permeability zones might be induced to the the oil zone, which allows large-amplitude oscillation of fluid
neighboring high-permeability zones by vibration-induced pressure or rock displacements.
crossflow. A simple two-layer reservoir model is employed to
show that vibration generates pore pressure waves that Laboratory Experiments for Improved Oil Recovery
propagate at different speeds in the low- and high- While there have been numerous laboratory experiments
permeability layers, thus creating a transient pressure which showed significant oil recovery by vibration, most were
differential across the layers and enhancing crossflow. Some done with idealized porous media such as very high-
example calculations are presented. permeability sandpack, and the vibration was usually applied
In the “Discussions and Conclusions” section, one by mounting the oil recovery set-up on an oscillating vibration
important topic that is generally not appreciated with the table with unrealistically large vibration amplitudes that are
seismics-based IOR is first discussed: the effective delivery of difficult to achieve practically in oil reservoirs.
vibrational energy to a large volume of reservoir that contains Simkin and Surguchev (1991) carried out a series of
oil. Unlike the use of seismic reflection for geologic imaging gravity segregation experiments, oil-displacement tests, and
for which the wave amplitude is not as important as the imbibition tests. In the gravity segregation tests with a
frequency composition, the IOR requires the energy supplied sandpack (3.5-cm diameter; 1-m long; φ = 0.33; k = 5.4 Darcy;
by the wave amplitude to displace the oil. Because of its Sor = 0.5; and oil viscosity = 1.65 cp), gravity equilibrium in
inevitable attenuation when a seismic wave propagates vertical position is established and, at t = 0, the pack is
through geologic formations, the vibrational energy actually inverted and the vertical distribution of oil saturation is
SPE 103870 3

measured with time. New gravity equilibrium is achieved at fingering suppression, nor increased oil mobilization, by
~12 days without vibration; but new gravity equilibrium was pusing.
achieved at ~2 hours with vibration of 120 Hz with amplitude Westermark et al. (2001) also reported some oil-
of 0.01 cm. In the oil-displacement tests in sandpack (35-cm x displacement test results, which showed that vibration has a
95-cm x 1-cm; φ = 0.45; k = 18-20 Darcy), 75% and 45% oil very slight increase in oil recovery. Single-phase permeability
recovery was observed with and without vibration, showed a sizable increase at a particular frequency, but they
respectively. In the imbibition tests in sandstone core (3-cm could not come up with an appropriate explanation.
D; 26-cm L; k = 1 Darcy), 56% oil recovery after 325 hours On the other hand, the experiments by the
without vibration, and 96% oil recovery with vibration was Waterloo/Alberta group (Wang et al. 1998; Spanos et al.
observed. Ashchepkov (1989) carried out a series of gravity 2003) showed significant effect of pressure oscillation at a
segregation experiments with sandstone cores of k = 0.28, very low frequency. A vertical column of sandpack (porosity
0.35, 0.65 Darcy (1-cm D; 60-cm L), to which vibration of 30, ~34%) was prepared and an oil (with viscosity 25.5 cp) was
60, 100, 200, 400 Hz frequency and 0.6 and 4.0 μm amplitude flown through it by establishing a hydrostatic pressure head at
was applied. Significantly faster imbibition of water and oil the top of sandpack. Pressure pulses were applied at the inlet
drainage was observed. Snarsky (1982) carried out a series of to sandpack at a frequency of ~1 Hz. A significant increase in
gravity segregation tests with glass tubes filled with fine flow rate was observed. To explain the permeability increase,
quartz sand (φ = 0.32-0.34; k = 9-11 Darcy). Vibration of 9 to a theory has been proposed (Spanos 2002, and references cited
40 Hz frequency and 0.2 W/cm2 power was applied. Good therein); but its physical basis has not been clearly spelled out.
drainage at 9-10 Hz was observed; drainage decreases around The possibility of employing vibration to remove NAPL
28-32 Hz, and then suddenly increases above 36 Hz. from contaminated soils has also been studied. Reddi and
Employing a synthetic core made from quartz sand and Challa (1994) put a water-saturated soil column on a vibration
marshallite (φ = 0.33; k = 5.4 Darcy), Pogosyan et al. (1989) table, and applied mechanical vibration at 60 Hz and with
investigated the effects of vibration on the single-phase amplitude ranging between 0 and 600 μm. They observed a
permeability and the gravity drainage. Vibration increased the significant increase in removal of NAPL. A mechanistic
permeability and enhanced the drainage; but the nature of the interpretation of the results is attempted by Reddi and Wu
vibration was not specified. (1996).
Ma et al. (1996) investigated the effects of vibration on As an application of their work (see above), Roberts et al.
core permeability (32 cores with 2.5-cm D; 3 to 7.5-cm L), (2001) attempted to remove DNAPL such as trichloroethylene
when translational vibration of 0.05-0.2 g force with (TCE) from a sandpack (of 110 Darcy; 1-in. D and up to 24-
frequency of 0-30 Hz was employed. The permeability is in. L) by applying vibration of 10-100 Hz frequency and
reported to increase by 28% when the frequency was core’s displacement of 1-40 μm. For vibration, a Terfenol-D
eigen frequency, which however was not defined. For other magnetostrictive actuator was used. TCE effluent
frequencies, the permeability actually decreased. Jin et al. concentration increased consistently with vibration
(1997) carried out oil-displacement experiments in 84 application.
different cores, and observed that vibration at the optimal The recent micromodel experiments carried out by Li et al.
frequency of around 200-300 Hz improved water injectivity (2005) also showed that vibration can mobilize trapped TCE.
and also improved oil recovery by 5-10%. Vibration was said They also observed that the mobilization is most effective for
to be most effective with low-permeability cores. a certain range of vibration acceleration amplitude.
Experiments carried out at Los Alamos National The overall conclusion that could be drawn from the above
Laboratory (Roberts et al. 2001; Roberts et al. 2003) are survey of the available laboratory oil recovery results is that,
notable in that, unlike Russian work, a compressional for consolidated porous media with permeability that is
vibration was applied to a sandstone core (of 50 md; 1-in. D usually observed in oil reservoirs (e.g., sandstone of ~100 md),
and 12-in. L) in its axial direction. An actuator delivering any significant mobilization of the waterflood residual oil by
dynamic strain as high as 10-4 induced pore pressure mechanical vibrations of realistically achievable amplitudes is
fluctuation of 0.3 psi amplitude. Its effects on permeability unlikely.
and oil/water relative permeabilities, and mobilization of For the unconsolidated systems, as briefly noted above, the
waterflood residual oil, were then studied for frequencies vibration seems to be more effective in mobilizing the non-
ranging from 20 to 1000 Hz. Some enhancement of oil flow wetting phase, but the reasons for the difference between the
was observed for frequencies between 25 to 100 Hz. consolidated and unconsolidated rocks are not known. Huh et
However, no significant changes in permeability have been al. (2004) observed from their sandpack experiments that the
observed, and no mobilization of residual oil occurred. gravity segregation between gas and oil occurs much faster
The above finding is similar to those made at AEA when vibration is applied, as observed by others earlier.
Technology, UK (Cable et al. 2001), except that a slight
increase in brine-phase mobility, rather than oil-phase Oil Recovery Mechanisms Proposed in Literature
mobility, was observed with pulsing. Here, over 60 oil- Quite a few mechanisms have been proposed earlier to
displacement tests with a sandstone core of 100 md, and a explain the improved oil recovery by vibration, but most of
sandpack of 13.5 Darcy were carried out, with imposed fluid them are not substantiated by experimental data or consistent
pressure pulsing at 0.2, 0.4, 1.0 and 2.0 Hz. γ–ray in-situ theoretical modeling work. Some of the proposed
saturation monitoring showed no evidence of viscous mechanisms are discussed here briefly. It has been recognized
early on that, if the improved oil recovery is from mobilization
4 SPE 103870

of oil ganglia from rock pores, vibrations with wavelength on and frequency. A similar numerical computation, with fluid
the order of pore length scale is required, either to create a properties and tube dimensions that are more realistic for
local pressure gradient in the neighborhood of the trapped oil actual residual oil mobilization, has also been carried out by
droplet, or to create a fluctuation of oil/water interface. Beresnev et al. (2005).
Because vibrations with such short wavelengths Because the seismic wave that reaches the oil zone will be
(consequently, high frequencies) do not travel a long distance of very small amplitude, and because a fluid pressure
in reservoir, and because the surface-generated seismic oscillation or rock displacement oscillation of a large
vibrations which are able to reach the oil zone will be of very amplitude would be required to mobilize oil, the possibility of
low frequency, an as-yet unknown link between low- resonance generation at the pore level has been investigated by
frequency vibration and oil recovery had to be found. Hilpert et al. (1997), Hilpert and Miller (1999), Hilpert et al.
Nikolaevskiy and coworkers (Barabanov et al. 1987; (2000), Iassonov and Beresnev (2003), and Beresnev et al.
Nikolaevskiy 1989; 1992a,b; Krylov et al. 1991; Lopuchov (2005). These works were based on the earlier observations
1999) proposed that certain low-frequency waves can cause by Dimon et al. (1988), Charlaix et al. (1988), and Charlaix
rock grains of the reservoir formation to generate high- and Gayvallet (1992) that, when subjected to an external
frequency waves. They suggested that the microscopic flow pressure oscillation, a liquid meniscus in a circular tube may
field generated by high-frequency waves may mobilize the show resonant oscillation, due to the interplay of surface
trapped oil droplets and also alter the oil/water relative tension, liquid inertia and liquid viscosity. In Hilpert et al.
permeabilities. While a non-linear theory has been proposed (2000)’s modeling study, the response of a residual oil ganglia
to advance the above concept (Beresnev and Nikolaevskiy subjected to a pressure oscillation was investigated in circular
1991), no experimental verification is yet available. tubes whose cross-section either is uniform or varies
Assuming that the high-frequency vibrations are locally sinusoidally. The resonance frequency was calculated as a
generated from rock grain movements, a number of function of the sand grain diameter of spherical packing that
speculative mechanisms have been proposed for mobilization serves as a model porous medium. They noted that the
of oil ganglia. Kuznetsov and Simkin (1991) suggested that resonance generation is not possible for small sphere
oil/water interfacial tension is reduced by two orders of diameters that are characteristic of oil reservoir rocks.
magnitude when subjected to vibration, thus mobilizing oil
ganglia, which subsequently coalesce with each other. PRELIMINARY EVALUATION OF OIL RECOVERY
Suggestion has also been made that vibration breaks up MECHANISMS
trapped oil ganglia into microscopic droplets of sizes much As reviewed above, the field evidences show that some
smaller than pore sizes, which are dispersed in water and flow incremental oil could be recovered from low-frequency
with water. Still another mechanism suggested is that seismic vibrations, but the mechanisms for recovery have not
vibration alters wettability making reservoir more oil-wet, thus yet been clearly identified. In this section, approximate
creating thin oil films spreading on rock surface (Kuznetsov models are developed to show that mobilization of waterflood
and Simkin 1991). residual oil by seismic vibration is unlikely, and that any
Another mechanism proposed by Ganiev et al. (1986, incremental oil by seismic vibration is likely from an
1989) is the pore fluid flow generated by periodic, transverse improved sweep of the oil bypassed by earlier recovery
deformation of pore wall. Assuming that representative pore methods, due to reservoir heterogeneity.
geometry is a circular tube, they showed that small
deformation of tube wall in the manner of traveling waves Mobilization of Waterflood Residual Oil by Vibration
could create a significant uni-directional flow in the tube, As described in the previous section, Hilpert et al. (2000)
presumably sufficient enough to mobilize an oil droplet derived an approximate expression relating the pore pressure
trapped at a pore throat. While physically more reasonable fluctuation with the volumetric displacement of the oil droplet
than the above hypothesis of high-frequency noise generation, in a pore. When seismic vibration is generated at the ground
the pressure gradient generated by rock deformation is in surface or at the wellbore, the vibrational energy is transmitted
general insufficient to mobilize the waterflood residual oil, as to the reservoir zone by displacements of fluids-filled rock
discussed in the next section. In a similar vein, Aarts and matrix as a poro-elastic medium. The pore pressure
Ooms (1998) and Aarts et al. (1998) calculated the net flow fluctuation is the consequence of the rock deformation,
generated when an ultrasonic wave propagates through a because the fluid pressure oscillation in rigid rock damps
porous medium. Because the ultrasonic vibration can down very quickly. Therefore, we investigate the feasibility
penetrated into porous media only for a short distance (~ of residual oil mobilization in terms of rock displacements.
cm’s), the theory was intended for applications such as near- In investigating the mobilization feasibility, we will utilize
wellbore cleaning (Roberts et al. 2000). the capillary number correlation that relates the mobilization
A detailed numerical calculation of full Navier-Stokes of residual oil to the capillary number, C a = μ w U γ o w ,
equations to investigate the effects of imposed fluid pressure where μw is water viscosity, U is interstitial velocity, and γow is
oscillations, on the movement of an immiscible fluid drop in oil/water interfacial tension. The correlation, shown as Figure
circular tubes of straight and sinusoidal longitudinal variation, 1 (from Delshad 1990), has been well established with many
has been made by Graham and Higdon (2000a,b). They careful oil-displacement coreflood experiments (e.g., Pope and
showed that, for fluid drops trapped at narrow constrictions Baviere 1991). Because the typical values of μw and γow in the
below the threshold pressure gradient, the oscillatory forcing capillary number correlation are known, we can estimate the
can mobilize the fluid drops for a certain range of amplitude
SPE 103870 5

fraction of the residual oil that is mobilized for a given and ε, for the tube radius of r = 100 μm. We see that even for
average interstitial velocity. the fairly large pore radius, a significant rock displacement is
When a seismic wave propagates through a reservoir, the required to mobilize even a fraction of the residual oil.
shear wave causes translational oscillations of rock pores, Figures 3 and 4 show So/Sori as a function of ω and ε, for r =
making the fluids in them slosh. The compressional wave 200 μm and 400 μm, respectively. For these tube radii which
causes oscillatory sqeezing of rock pores, making the fluids in represent very coarse sands or laboratory beads pack, a
them again slosh. Compressional wave would, however, significant portion of the residual oil could be mobilized if the
squeeze the pore throat more than the pore body, so that it will rock displacement is large. We note, however, that achieving
probably push an oil ganglion away from the throat, instead of such a large rock displacement at a sufficiently high
forcing it through the throat. In fact, the “squirt flow” frequency, at the reservoir zone, would be normally quite
mechanism of squeezing fluid from thin gaps has been difficult. It is interesting to note that for the micromodel
recognized as an important mechanism for seismic velocity experiments carried out by Li et al. (2005), the pore throat
attenuation (Bourbie et al. 1987; Li et al. 2001; Miksis 1988). radius was 380 μm, which is comparable to the conditions of
Assuming accordingly that the translational oscillation Figure 4. While the residual oil mobilization from very
produces more fluid displacements, and that the pore space in coarse-grained systems with large-amplitude translational
reservoir rock consists of cylindrical tubes (of same radius, r) wave has been demonstrated in the laboratory experiments, as
of random orientation, we now calculate the average velocity reviewed earlier, the above approximate model study suggests
of a fluid which sloshes in the tubes. When a tube makes a that the residual oil mobilization from actual reservoirs would
translational periodic displacement, in parallel to tube axis, be unlikely.
iω t
according to u s = ε e (where ε is the displacement As described earlier, the laboratory experiments carried out
by Roberts et al. (2003) and Cable et al. (2001) confirm this
amplitude and ω is the wave frequency), the average fluid assessment.
velocity in it can be calculated as (Bedford et al. 1984)
Recovery of By-Passed Oil by Vibration-Induced
2ε J 1 (ω r β ) [1] Crossflow
Uz =
⎛ ωr ⎞ During waterflood in a heterogeneous reservoir, the
⎜ ⎟ J 0 (ω r β )
⎝ β ⎠ injected water displaces oil from high-permeability layers
much more effectively than from low-permeability layers,
iωμw leaving behind substantial amounts of oil in the latter. In this
where β 2 ≡ ; and ρw is water density; and J0 and J1 are subsection, an approximate theoretical analysis is made to
ρw estimate the extent with which the vibration-induced crossflow
the zero and first order Bessel function of the first kind. When across the heterogeneous reservoir layers extracts the bypassed
a tube makes a translational periodic displacement oil from the low-permeability layer. While it is quite difficult
perpendicular to tube axis, there will be no net flow in the to bring the injected water to the low-permeability zones in a
direction of tube axis (Bedford et al. 1984). For a tube whose highly heterogeneous reservoir, vibrational wave can transmit
axis makes an angle of ψ with the direction of seismic wave, effectively through both high- and low-permeability zones.
therefore, the average displacement of fluid in its axis When a seismic vibration propagates through a heterogeneous
direction is reservoir, the pore pressure response in different permeability
layers becomes different, generating a transient pressure
U = U z cos ψ [2] difference between the layers. The pressure difference
generates crossflow, potentially sweeping out oil from the
Because the tubes in the rock are assumed to have all low-permeability zone.
orientations with equal probability, the capillary number will In the seismic velocity attenuation studies, importance of
be first calculated with Eq.[2] for tubes of ψ-orientation; the crossflow has been recognized by White et al. (1975), Norris
residual oil recovery efficiency is then calculated from the Ca (1993) and others. In White et al.’s simplified treatment
correlation (Fig. 1); and the overall recovery efficiency is applicable at low-frequency limit, a temporally-periodic
obtained by integrating over all orientation angles. Instead of variation of normal load is applied at the top and bottom
the detailed, actual oscillatory response, the modulus of the boundaries of reservoir which consists of layers alternately
average velocity is employed for the above calculation, which filled with gas and liquid. As the rock stress is transmitted
may be viewed approximately as the maximum interstitial vertically through the layers, fluid pressure in gas- and liquid-
velocity available when the translational displacement of rock filled layers becomes different because of the large difference
due to shear wave is ε. in compressibility, and crossflow of fluids occurs between the
For typical values of γow = 20 dynes/cm, μw = 1 cp, and ρw layers. All investigations of seismically-induced crossflow
= 1 g/cm3, the residual oil mobilization efficiency was appear to consider only flow of single-phase fluid (Bourbie et
calculated for a range of tube radius (r), frequency (ω) and al. 1987). How the seismically-generated pressure gradient
affects the oil displacement behavior, with accounting of
rock displacement (ε). For tube radius smaller than 50 μm, no
multi-phase mobility effects, has not been investigated before,
residual oil mobilization was observed for ω = 1 – 200 Hz and
and is carried out here approximately.
ε = 1–100 μm. Figure 2 shows So/Sori (oil still remaining after
treatment/ residual oil initially in the rock) as a function of ω
6 SPE 103870

In addition to the crossflow induced by rock stress iω D

propagation, crossflow can also occur from propagation of A1iω pa1 − σ a − T ( pa1 − pa 2 ) = 0 [5]
M1 h1
fluid pressure waves. In well testing, crossflow among the
layers of different permeability is known to occur, making the
interpretation of the test results difficult (Gao and Deans iω D
A2 iω pa 2 − σ a − T ( pa 2 − pa1 ) = 0 [6]
1988). When the pressure at the wellbore is suddenly M2 h2
changed, the pressure shock front travels in different layers of
the reservoir at different speeds, due to the difference in By re-arranging Eqs.[5] and [6], we obtain the difference
permeability and in rock and fluid compressibilities. In the in pressure oscillation amplitudes:
areas where the pressure shock front already passed through in
one layer but did not in the neighboring layer, a pressure A2 A
gradient thus develops between the layers temporarily, causing − 1
pa1 − pa 2 M1 M 2
crossflow. Unless the frequency of the pressure oscillation is = [7]
ultralow, however, this purely fluid pressure-based effect is σa D ⎛A A ⎞
A1 A2 − T ⎜ 1 + 2 ⎟
known to be small, because the pressure oscillation in rigid iω ⎝ h2 h1 ⎠
porous media generally dies down very quickly. In the present
approximate analysis, therefore, only the rock stress-induced The transient pressure differential between the two layers,
crossflow will be considered. in response to the imposed normal stress oscillation, σ, can
To assess approximately the extent of crossflow, a simple now be estimated in terms of the permeabilities, fluid
two-layer reservoir model is considered (see Figure 5), saturations and compressibilities and rock modulus. Again,
employing a set of coupled equations governing rock the modulus of the pressure differential (scaled with the
deformation due to seismic vibration and the resulting fluid imposed normal stress amplitude) from Eq.[7] is calculated.
flow. The derivation of the equations is given in Appendix III.
For the sample calculations, the following parameter
The equations are solved here for a simple case to investigate
values were employed, unless otherwise specified: μw = 1 cp;
the possibility of improved sweep of the bypassed oil. The
simplified case we consider is when the high-permeability
μo = 2 cp; φ1 = φ2 = 0.25; h1 = h2 = 10 m; Kw = 3.3 x 105 psi;
Ko = 1.0 x 105 psi; M = 6.09 x 105 psi (White 1975; White et
layer is at near the waterflood residual oil saturation (Sorw)
al. 1975). The oil/water relative permeabilities were
condition and the low-permeability layer still contains oil at a
saturation close to the initial saturation which is much higher interpolated from the following array:
than Sorw. We assume for this simple case that there is no
Sw = 0.10, 0.25, 0.40, 0.55, 0.60, 0.72,
injection of water (thus, the reservoir is initially in a static
condition) and Soi (i = 1, 2) are constant. As discussed in the kro = 1.0, 0.27, 0.06, 0.0043, 0.0010, 0.0,
Appendix, we assume that the seismic vibration is delivered to krw = 0.0, 0.005, 0.022, 0.080, 0.125, 0.35.
the reservoir zone in the vertical direction only, uniformly in
x-direction (similar to the White et al. case). The normal For the transmissibility parameters for crossflow between the
two layers, DT and Do defined as Eqs.[A15a,b], we simply
stress term in Eqs.[A16] and [A17] in Appendix III therefore
assumed that there is no skin effect, i.e., hv = 0. Because our
serves as a forcing term that causes flow. Neglecting the
simplified model involves no horizontal flow, the
pressure gradient in the x-direction in Eq.[A16], we obtain
permeabilities (k1 and k2) represent the layer’s vertical
permeability. It is emphasized that, since the representation of
∂ p1 1 ∂σ DT the crossflow with DT and Do is a gross simplification, the
A1 − − (p − p ) = 0 [3]
∂ t M1 ∂ t h1 1 2 predictions of the pressure differential and the resultant oil
transfer rate, given below, are only qualitative.
∂ p2 1 ∂σ DT Figure 6 shows the modulus of the pressure differential
A2 − − (p − p ) = 0 [4] (scaled with the imposed normal stress amplitude) as a
∂ t M 2 ∂ t h2 2 1
function of the frequency (ω) for different combinations of the
two-layer permeabilities (in md’s). The layer oil saturations
The definition of the terms in the above equations, and were assumed to be So1 = 0.35 and So2 = 0.60. The fluid
those that follow, are given in Appendix III, and also are listed pressure differential generated in reponse to the imposed
in Nomenclature. normal rock stress is seen to be modest. Figure 7 shows a
If the normal stress forcing function is periodic with similar case except that the layer oil saturations are now So1 =
amplitude of σa and frequency of ω, as shown below, we 0.30 and So2 = 0.80. With a larger difference in fluid
assume that the layer pressures also vary periodic in time: saturations, we see that the pressure differential increases
substantially. Figure 8 is similar to the case of Figure 6,
σ = σ a e iω t ; p j = paj eiωt except that oil is more compressible than before (Ko = 0.75 x
105 psi instead of 1.0 x 105 psi). We see that the pressure
differential increases somewhat with an increased difference
where j = 1, 2 are for high- and low-permeability layers, in fluid properties. In the above, we simply assumed that M 1
respectively; and paj is amplitude of pressure oscillation, = M 2, even though they would be different with differences in
which arises in response to the rock stress oscillation. Eqs.[3] permeabilities and fluid saturations. Figure 9 is similar to the
and [4] then become
SPE 103870 7

case of Figure 6, except that the high-k layer is more DISCUSSIONS AND CONCLUSIONS
compressible than the low-k layer (M1 = 5.0 x 105 psi instead In the above section, the difficulty of mobilizing
of 6.09 x 105 psi). We again see that the pressure differential waterflood residual oil was shown employing the well-
increases substantially. established capillary number correlation and a simplified pore-
The rate of oil saturation change arising from the crossflow level model. The capillary force holding oil ganglia in pore
across the two layers will be approximately throats is much too strong for a realistically achievable
vibration amplitude to overcome. In order to explain the
∂So1 ⎡ Do ⎤ improved oil recovery sometimes observed, it was then shown
φ1 =⎢ ( pa1 − pa 2 ) ⎥ eiωt [8] that the oil bypassed by earlier recovery methods, and still left
∂t h
⎣ 1 ⎦ in the low-permeability zones of a heterogeneous reservoir,
might be induced to the neighboring high-permeability zones
Inserting Eq.[7] into [8], we obtain by vibration-induced crossflow. This mechanism of improved
recovery of bypassed, but still mobile, oil may explain why
⎡ iDo A2 A ⎤ the success of the seismic vibration is rather erratic. Because
⎢ ( − 1 ) ⎥ of the stochastic nature of heterogeneity, the accurate
∂So1 ⎛ σ ω ⎞ h1 M1 M 2
= ⎜ a ⎟⎢ ⎥ eiωt [9] characterization of the locations and amounts of the bypassed
∂t φ ⎢ ⎞⎥
⎝ 1 ⎠ ⎢ iω A A − D ⎛ A1 + A2 oil is extremely difficult, and consequently, an accurate
1 2 T ⎜ ⎟⎥ estimation of the incremental oil from vibration application is
⎢⎣ ⎝ h2 h1 ⎠ ⎥⎦
also very difficult.
If seismic vibration could indeed enhance crossflow
In this qualitative estimation of the instantaneous oil
between the high- and low-permeability zones, one potential
saturation change due to crossflow across the layers, we note
application is its use for enhanced delivery of IOR chemicals
that the saturation change will not follow the periodic
to the low-permeability zones, which otherwise would be
oscillation of the pressure difference between the two layers.
difficult. For example, the use of a low concentration of
We assume approximately that when p1 – p2 < 0, a small
surfactant has been shown to be a promising way of inducing
amount of the bypassed oil in the low-k layer (2) is pushed to
oil out of the matrices of naturally fractured reservoirs, by way
the already well-swept high-k layer (1); but when p1 – p2 > 0,
of wettability alteration (Hirasaki and Zhang 2004; Seethepalli
the oil does not move from the high-k to low-k layer because
et al. 2004). Seismic vibration may be an effective way of
still So1 < So2. With this consideration, we use one-fourth of
squeezing the injected surfactant into the matrix from fracture,
the modulus calculated from Eq.[9] as a qualitative measure of
instead of relying only on diffusion and dispersion. Such
the instantaneous oil transfer rate (scaled with the imposed
application potential warrants some detailed consideration.
normal stress amplitude) from the low-k to high-k layer.
For the sample calculations, we employed the parameters
Vibrational Energy Delivery
that were used for the above pressure differential calculation,
In the above, the possible mechanisms for oil recovery by
unless otherwise specified. We also assumed k1 = 1 md and k2
vibration application were investigated. An equally important
= 0.01 md. Figure 10 shows the instantaneous oil transfer rate
question that needs to be addressed is how the vibration
(scaled with the imposed normal stress amplitude) as a
energy is delivered effectively from its source to the reservoir
function of the frequency (ω) and different combinations of zone. This is important because, as shown above, to sweep
the layer oil saturations. The different oil saturation out the oil in the low-permeability zone in any meaningful
combinations are meant to crudely imitate the oil saturation amounts, a continuous application of vibration with a
changes in the two layers as the crossflow progresses. The significant amplitude is needed. For the vibration generated at
rate of oil saturation change is initially quite high, probably the ground surface employing vibroseis, its amplitude is
because the effective crossflow transmissibility (Do) is generally known to be attenuated rapidly as it propagates
overestimated. As So1 and So2 approach closer to each other, through the near-surface alluvium and the overburden, before
the transfer rate decreases sharply. The qualitative calculation it reaches the reservoir layers (Bourbie et al. 1987).
suggests that the oil transfer due to the transient pressure Generation of vibration inside the wellbore at the reservoir
differential may be possible. Figure 11 is similar to the case zone, that has a significant rock displacement amplitude, is
of Figure 10, except that oil is more compressible than before difficult due to the space limitations. Energy delivery into the
(Ko = 0.75 x 105 psi instead of 1.0 x 105 psi). We see that the reservoir zone requires rock deformation wave, because any
rate of oil saturation change increases somewhat with an fluid pressure oscillation with frequency higher than even a
increased difference in fluid properties. Figure 12 is similar tenth of a Herz damps quickly.
to the case of Figure 10, except that the high-k layer is more For the above reasons, only meaningful way of effectively
compressible than the low-k layer (M1 = 5.0 x 105 psi instead delivering vibrational energy appears to be the generation of
of 6.09 x 105 psi). We see that the rate of oil saturation resonance in the reservoir zone. With resonance, a continuous
change increases substantially. The above qualitative model delivery of a small-amplitude vibration could be excited to a
calculations are only meant to show that the recovery of the large-amplitude rock displacement at the reservoir zone. For a
bypassed oil by vibration is a possibility, and a more detailed, qualitative look at the resonance possibility, let Lr be the
quantitative model prediction is warranted. characteristic thickness of an oil reservoir, which has a
sufficient wave reflection capacity against its overburden and
underburden; and Ur be the seismic wave velocity in the
8 SPE 103870

reservoir. A resonance frequency would be ωr ~ Ur/Lr. For Conclusions

typical compressional and shear wave velocities (in fluid- The following conclusions are made based on the
saturated rock) of 2.2 x 105 cm/sec and 1.0 x 105 cm/sec, literature review and the approximate modeling evaluation of
respectively (White 1975), the resonance frequency would be the possible mechanisms for the oil recovery:
ωr ~ 220 and 100 Hz, if Lr = 10 m. If we recognize the fine • A critical review of the vibration experiments and the
layering of reservoir and assume Lr = 1 m, ωr ~ 2200 and 1000 modeling studies reported in literature suggests that the
Hz, which are much higher than the frequencies that are mobilization of waterflood residual oil by seismic
reported to be effective for oil recovery (1 ~ 40 Hz). This vibration is not likely, for consolidated porous media with
qualitative argument suggests the possibility of a much slower permeabilities normally observed in oil reservoirs
wave velocity. (10~1000 md’s). The capillary force holding oil ganglia
The possible role of a slow wave for resonance generation in pore is much too large to overcome by pressure
has been proposed by Bachrach et al. (1998). In order to gradient or acceleration normally achievable by seismic
explain the liquefaction of sand layer after an earthquake, they vibration.
showed that the resonant amplification of Biot wave could • An approximate application of a simple pore-level model
drastically alter the poroelasticity of the sand layer, leading to to the well-established capillary number correlation
its liquefaction. It is worth to note that the extension of the provides residual oil mobilization predictions that are in
Biot formulation (Biot 1956a,b; Bourbie et al. 1987) to agreement with the above assessment. With the
account for the presence of the two immiscible fluid phases in vibrational amplitudes that could be realistically achieved
poroelastic media predicts the possibility of two slow waves. in the reservoir, the amount of residual oil mobilized
Employing mixture theory to derive macroscopic equations for would be negligible.
wave propagation in porous media with two immiscible fluids, • The improved oil recovery that results from vibration
Santos et al. (1990a,b; 2004) predicted that two slow waves application in the field is likely from the improved sweep
may arise in addition to the normal compressional and shear of the oil in the low-permeability zones that was bypassed
waves. Employing volume-averaging technique, Tuncay and by earlier recovery techniques. When a seismic vibration
Corapcioglu (1996; 1997) also extended the Biot equations to propagates through a heterogeneous reservoir, the pore
account for two immiscible fluids and predicted that the two pressure response in different permeability layers
additional waves may have velocities much slower than even becomes different, generating a transient pressure
the Biot wave. Recently, Lo and Sposito (2002) and Lo et al. difference and crossflow between the layers, potentially
(2005) also derived the two-immiscible fluids extension of squeezing out oil from the low-permeability zones. An
Biot equations, employing the continuum mixture theory. If approximate model calculation employing a simple two-
such slow waves due to the presence of two immiscible fluids layer reservoir model suggests that the above may be a
exist in any meaningful magnitude and the resonance is indeed viable mechanism.
generated from such waves, the vibration generation at the • A continuous feed of seismic wave to the reservoir zone,
source at resonance frequency will be an effective way of and subsequent development of resonance there, appears
delivering the vibration energy for a long distance. This to be a way to generate large-amplitude rock displacement
possibility needs to be further considered. and fluid pressure oscillations, which may in turn squeeze
While the above extensions of the Biot equations lead to oil out of the low-permeability zones of a heterogeneous
the potentially important possibility of resonance reservoir. To develop resonance for reservoir length scale
development, one weakness of deriving the macroscopic (in m’s), the presence of a slow wave is needed, and the
equations by use of mixture theory or volume-averaging Biot slow wave or its extension to immiscible fluid phases
method is that they do not provide direct linkage to the pore- offers potential for resonance quantification, given the
level mechanisms. In this regard, the detailed laboratory and reservoir dimensions and its fluid and rock properties.
modeling work carried out to understand the role of the
fluctuations of interface and contact line formed by
immiscible fluids, on seismic velocity attenuation, is worth a Appendix I: Field Evidences for Improved Oil
brief description here. The subject has relevance to the Recovery
seismics-based IOR. Miksis (1988) was first to recognize the There is a sizeable body of literature, mainly from Russia,
potential importance of the contact line fluctuations for on the field application of seismic vibration. As mentioned
seismic velocity attenuation, from model calculations earlier, however, no definitive demonstration of the vibration’s
employing a 2-D bubble imbedded in a crack. Moerig et al. effectiveness is available. Some of the representative results
(1997), Schutt et al. (2000), and Brunner and Spetzler (2002) are briefly discussed below.
carried out a detailed laboratory characterization of vibration Changirtash Field: Two field tests were carried out at
attenuation in micromodel cracks which have a well-defined Changirtash field in Kirgizstan, whose reservoir is at a shallow
fluid contact line. A comprehensive review of studies on depth of 240-448 m, with a thickness of 4-10 m and a dip of
pore-level oscillatory behaviors of immiscible fluids, and their 35°. Average permeability was 200 md, and oil viscosity was
possible connection to macroscopic seismic wave attenuation, 35 cp. For the first test, two 50 kW surface vibrators were
is given by Li et al. (2001). installed 20 m apart, and were vibrated from October 5 till
October 20 of 1988. The initial water cut of 69% was
decreased by 25-30%, and an incremental oil of 300 tons was
SPE 103870 9

recovered from the test area of 600 x 600 m. The second test response (Xue and Yang 1996). Zhang et al. (1999) gives a
was carried out at a deeper portion of the reservoir (511-708 somewhat more detailed description of a field test at Block
m), and vibration was applied for the month of December, 1764 in Xinjiang province, carried out in July, 1993. A 20-ton
1988. An incremental oil of 2540 tons was recovered from the off-center vibrator with maximum vibration force of 200 kN
test area of 1500 x 1500 m, which included 16 producers. was used to apply vibration for three periods (46, 48, and 40
Other details were not provided (Kouznetsov et al. 1998; hours). Until November of that year, 2230 tons of incremental
Kuznetsov et al. 2002). oil is said to have been produced in the test area of 0.4 km2. In
Jirnovskyi Field: Two tests were carried out at Jirnovskyi a recent paper, Zhu et al. (2005) described the development of
field in Volgogradskiy region of Russia. Reservoir depth is a downhole harmonic vibration oil-displacement system
about 1000 m, and permeability varies widely in the range of (DHVOS) by China National Petroleum Co. researchers,
0.1-5 Darcy. Oil viscosity is 4.5 cp. For the first test, which generates vibration of 88 shocks per minute with each
vibration was applied from October 20 to December 15, 1991. shock having 5-ton strength. More than 100 oil-producing
Average water cut was 92%, which was decreased by 6%. For wells have been stimulated since 1997, at their Liaohe, Huabei
the second test, vibration was applied from May 18 to July 13, and Shengli fields, with significant increase in oil cut and oil
1992. Average water/oil ratio decreased from 9.7 to 8.3. A production. The details of the oil performance data are
notable feature for these tests was that a specially designed however not available.
wave-transmitting system was lowered in the wellbore to the Morgan Field: A Canadian service company (PE-Tech,
pay zone, to improve vibration delivery efficiency. The Inc.) had been conducting a field test of the concept advanced
detailed nature of the wave guide is not known (Kouznetsov et by the researchers of Universities of Alberta and Waterloo.
al. 1998; Kuznetsov et al. 2002). The test, which started in December, 1998, is carried out in the
Pravdinsk Field: The test, which applied low-frequency Sparky Sands of Morgan field, Alberta, in cooperation with
seismic vibration from the ground surface using a heavy-duty the field's owner, Wascana Energy. In the test, a modified rod
electromagnetic vibration generator, was carried out at a pump with a plunger generates fluid pressure pulses at the
watered-out section in the southern part of the Pravdinsk field wellbore, at a rate of 3 strokes per minute. By the middle of
in Siberia. The reservoir, whose depth is 2340 m, has a pay February 1999, the field showed a 67% production rate
thickness of 8.1 m, porosity of 20%, permeability of 104 - 117 increase in the four offset wells and a 37% increase on average
md, and oil viscosity of 1.35 cp. The particular reservoir for all wells in the section. The size of the test area and the
section had 39 producers and 27 injectors, but the oil duration of pulse application are not known (Dusseault 1993;
production analysis was limited to 18 producers which were Spanos et al. 2003).
within a 3-km radius from the vibration generator. While Elk Hills and Lost Hills Field: At the diatomite reservoir
vibration was applied from Sept. 17 to Nov. 11, 1994, the oil in the Lost Hills field in California, the wellbore vibration was
production analysis was made for (a) 8.5 months before; (b) 2 generated for 12 days, and then for 38 days after two months.
months during; and (c) 7 months after the period of vibration The vibration service provider reported that, of the 80
application. Unfortunately, the exact details of the vibration producing wells within a half-mile radius of the vibration
applied are not known, even though the vibrator is reported to source well, 26 wells showed significant increase in oil cut
be able to generate vibrations with a force amplitude up to 10 and oil production after the first period of vibration
kN and frequency of 5 - 20 Hz range. Upon applying application. After the second vibration application, 48 wells
vibration, for a group of 13 wells, the average water cut was showed increase in oil cut and production. The wellbore
reduced from 91.9% to 83.3% while the average fluid vibrator was also applied at the Elk Hills field in California
production rate decreased slightly from 46.9 to 44.8 t/day. For (operated by Oxy), and increased oil cut and oil production
the remaining 5 wells, however, the average water cut were reported from 73 wells within a half-mile radius of the
increased from 85.1% to 88.1%, while the average production vibration source well. The seismic responses were also said to
rate increased from 30.6 to 36.4 t/day. The effect of the be measured by the Lawrence Berkeley National Laboratory
vibroseismic stimulation persisted for 7 to 12 months after the researchers at various locations during the seismic stimulation,
application (Simonov et al. 1996). but their interpretation has been reported (Kostrov et al. 2001;
Mortymya-Teterevsky Field: This is not an oil recovery Tech. Update, “Value of In-Situ Seismic Waves – Regain Lost
test, but a detailed field investigation of the seismic wave Reserves, Increase Oil Cut”, J. Petrol. Tech., 24-25, April
propagation and the frequency/amplitude responses carried out 2005).
by Kurlenya and Serdyukov (1999a,b) at Mortymya- A heavy oil field in Germany: The wellbore vibrator
Teterevsky oil deposits in Russia. Of particular note is their developed by the Canadian group (see Appendix II) was
detection of resonance frequency, which they attempted to employed for a half-year field test at a weakly consolidated,
relate to the reservoir’s lithology. heavy oil (90-120 cp) reservoir operated by Shell
Chinese Field Tests: Based on the Russian field success International. The water injection for waterflooding was
reports, the Jilin Oil Fields and the National Seismology combined with the downhole hydraulic pulsing at a frequency
Bureau began cooperative research on the vibration of 5-6 pulses per minute with a 4-17 bar power. There was a
technology during the late 1990. The Jilin, Liaohe, Daqing, definite improvement in the water injectivity, but whether the
and Yumen fields are said to have successfully carried out a stimulation enhanced oil production or not was difficult to
total of 15 field tests, but their details are not available. For judge (Groenenboom et al. 2003).
reservoir area of up to 1000 m radius, and depths ranging from
400 to 1290 m, 70% of wells is said to have shown a positive
10 SPE 103870

Appendix II: Vibrational Energy Delivery Hardwares low-frequency vibrations of our interest are large, the spatial
In this Appendix, the vibration generation hardwares variations of porosity and fluid densities are small. Eq.[A1]
reported in literature are briefly discussed. On above-ground then becomes
vibration generators, Russians have developed, field-tested,
and manufactured them in quantity, which can generate ∂φ
seismic waves of different amplitude and frequency − + ∇ ⋅ [ (1 − φ ) w] = 0 [A4]
∂t
(Kouznetsov et al. 1998; Kuznetsov et al. 2002; Simonov et
al. 1996).
On hydraulic pulse generation at the wellbore downhole, which, with the volume strain, ε ≡ ∇ ⋅ u , becomes
one notable example is the scheme offered by Applied Seismic
Research (which carries a Russian technology license, Vagin ∂φ ∂ε
1996), which uses a modified rod pump to inject ~5 barrel − + (1 − φ ) =0 [A5]
∂t ∂t
squirts of water into formation about 5 to 10 times per minute,
and generate pressure waves in a water-hammer fashion
Eq.[A2] can be rewritten
(Kostrov et al. 2001). A number of oil companies are believed
to be currently testing the tool in mature reservoirs in U.S., as
described above in the field activity review (see Elk Hills and ∂φ φ S w ∂ p ∂ Sw
Sw + +φ + ∇ ⋅ (φ S w v w ) = 0 [A6]
Lost Hills, above). Another wellbore, hydraulic pulse ∂ t Kw ∂ t ∂t
generation tool is offered by PE-TECH of Alberta, Canada,
whose mechanism appears to be very similar to the above tool where
(Dusseault 1993; Spanos et al. 2003). Etrema Products, d ρw dp d ρo dp
manufacturer for a magnetostrictive material, Terfenol-D = ; =
ρw Kw ρo Ko
(alloy of terbium, dysprosium and iron), has a vibrator with
two Terfenol rods (5-cm D; 25-cm L) which periodically
squeeze and relax a cylindrical rubber plug sandwiched With the Darcy’s equation
between them. Fluid pressure oscillation of ~2 psi amplitude
is generated at 250-400 Hz frequency. As briefly described in φ S w (v w − w) = −λw (∇p − ρ w g ∇D) [A7]
Appendix I, Zhu et al. (2005) reported on a downhole
harmonic vibrator that generates vibration of 88 shocks per and Eq.[A5] inserted into it, Eq.[A6] for water phase then
minute with each shock having 5-ton force. becomes
A host of different vibration generation methods have been
proposed, e.g., electro-acoustic method (Ellingsen 2002), and ∂ε φ S w ∂ p ∂ Sw
downhole asymmetric rotator (Westermark 2001, 2002); but (1 − φ ) S w + +φ − ∇ ⋅ ⎡⎣λw (∇p − ρ w g ∇D) ⎤⎦ = 0
∂ t Kw ∂ t ∂t
they will not be discussed here.
[A8]
Appendix III: Derivation of Reservoir Flow and Rock
Deformation Equations kkrw kk
where λw ≡ ; λo ≡ ro ; k is permeability; krw, kro
To describe rock deformation and fluid flow, the mass μw μo
conservation equations for rock, water and oil are first written: are water and oil relative permeabilities; μw, μo are water and
oil viscosities; g is gravity constant; and D is reservoir depth.
∂ For simplicity, the capillary pressure term is omitted in this
⎡(1 − φ ) ρ s ⎦⎤ + ∇ ⋅ ⎣⎡(1 − φ ) ρ s w⎦⎤ = 0 [A1]
∂t⎣ derivation.
A similar derivation can be made for the oil phase:
∂ (φρ w S w )
+ ∇ ⋅ (φρ w S w v w ) = 0 [A2] ∂ε φ So ∂ p ∂S
∂t (1 − φ ) So + + φ o − ∇ ⋅ ⎡⎣λo (∇p − ρo g ∇D) ⎤⎦ = 0
∂ t Ko ∂ t ∂t
∂ (φρo So ) [A9]
+ ∇ ⋅ (φρo So v o ) = 0 [A3]
∂t By summing up Eqs.[A8] and [A9], we obtain

where φ is porosity; ρs, ρw, ρo are densities of rock, water and ∂ε ⎛S S ⎞∂ p

oil, respectively; Sw, So are water and oil saturations; vw, vo are (1 − φ ) +φ ⎜ w + o ⎟
∂u
∂t ⎝ K w Ko ⎠ ∂ t [A10]
interstitial velocities of water and oil; and w = , u being
∂t − ∇ ⋅ ⎡⎣λT ∇p − (λw ρ w + λo ρo ) g ∇D ⎤⎦ = 0
rock displacement vector. The following assumptions are
made: (i) when a vibrational wave propagates, porosity change where λT = λw + λo.
occurs due to rock grain rearrangements while the rock grain The temporal change in volume strain, ε, in Eqs.[A8]
density remains constant; and (ii) since the wavelengths for to [A10] will be governed by rock deformation response as
SPE 103870 11

vibrational wave propagates in the reservoir. For rock ∂ pi Soi ∂σ ∂ Soi ∂ ⎡ ∂ pi ⎤ Do

Bi − + φi − λ − (p − pj) = 0
deformation, we consider the simplest case of a compressional ∂ t Mi ∂ t ∂ t ∂ x ⎢⎣ oi ∂ x ⎥⎦ hi i
wave propagating uniformly in vertical direction only. Such
will be the case when a seismic vibration is generated from [A17]
ground surface above the reservoir. We will then have only
the vertical, normal stress component, σzz = σ, and the volume where the subscript i designates the variable to be for the i
strain becomes layer.

Nomenclature
1
ε =− (σ − p) [A11] A, B Effective compressibilities defined by Eq. [A14a,b]
M Ca Capillary number
D Reservoir depth
where M ≡ (1 − φ ) ρ sV p2 is the p-wave modulus of the porous DT, Do Effective total (and oil) transmissibility across two
rock, and Vp is the p-wave velocity through the porous rock. layers, Eq. [A15]
The normal stress, σ, is viewed, in our simplified qualitative hi Thickness of reservoir layer (i=1, 2)
application, as an external forcing function, and is a function k Permeability
only of t. krw, kro Water and oil relative permeabilities
In (x, z) coordinates, Eqs.[A10] and [A9] can be Kw, Ko Water and oil bulk modulus
rewritten, respectively, M Rock modulus, Eq. [A11]
pi Fluid pressure in i-layer
∂ p 1 ∂σ ∂ ⎡ ∂ p ⎤ ∂ ⎡ ∂ p ⎤
A − − λ − λ = 0 [A12] r Tube radius
∂ t M ∂ t ∂ x ⎢⎣ T ∂ x ⎥⎦ ∂ z ⎢⎣ Tz ∂ z ⎥⎦ Swi, Soi Water and oil saturations in i-layer
t Time
∂ p So ∂σ ∂ So ∂ ⎡ ∂ p ⎤ ∂ ⎡ ∂ p ⎤ us Rock displacement vector
B − +φ − λ − λ =0 vw, vo, w Water and oil velocity; rock velocity
∂t M ∂t ∂ t ∂ x ⎢⎣ o ∂ x ⎥⎦ ∂ z ⎢⎣ oz ∂ z ⎥⎦
x, z Horizontal and vertical coordinates
β Parameter defined for Eq. [1]
[A13] ε Oscillatory rock displacement
φ Porosity
where γow Water/oil interfacial tension
⎛S S ⎞ 1 ⎛ φ 1 ⎞ λw, λo, λT Water and oil mobilities; total mobility
A≡φ⎜ w + o ⎟+ ; B ≡ So ⎜ + ⎟ [A14a,b] μw, μo Water and oil viscosities
⎝ K w Ko ⎠ M ⎝ Ko M ⎠
ρw, ρo Water and oil densities
σ Normal stress
We now consider a simple two-layer reservoir model,
ω Frequency
in which averaged values of pressure, phase saturations, and
fluids and rock properties are employed in each layer. The
References
crossflow between the layers is assumed to be linearly
1. Aarts, A. C. T., and Ooms, G., "Net Flow of
proportional to the pressure difference and empirical
Compressible Viscous Liquids Induced by Traveling
transmissibility constants, Do and DT. To estimate the extent
Waves in Porous Media", J. Eng. Mech., 34, 435-450
of crossflow during pressure testing at wells (Gao and Deans
(1998).
1988), a similar approximation has been employed with the
2. Aarts, A. C. T., Ooms, G., Bill, K. J., and Bot, E. T. G.,
following relations:
"Enahncement of Liquid Flow through a Porous Medium
2 2 by Ultrasonic Radiation", SPE 50594, presented at SPE
DT = ; Do = [A15a,b]
2hv h1 h2 2hv h1 h European Petrol. Conf., The Hague, The Netherlands,
+ + + + 2
λTv λT 1 λT 2 λov λo1 λo 2 Oct. 20-22, 1998.
3. Ariadji, T., "Effects of Vibration on Rock and Fluid
where hi is the i-layer thickness, and hv is the thickness of thin Properties: On Seeking the Vibroseismic Technology
skin which may exist between the two layers; λTi and λoi are Mechanisms", SPE 93112, presented at SPE Asia Pacific
the i-layer total mobility and oil-phase mobility, respectively; Oil Gas Conf., Jakarta, Indonesia, Apr. 5-7, 2005.
and λTv and λov are the total and oil-phase mobility of the thin 4. Ashchepkov, Yu. S., “Infiltration Characteristics of
skin. Because hv, λTv and λov are in reality not known, they are Inhomogeneous Porous Media in a Seismic Field”,
used as adjustable parameters. Fiziko-Tekhnicheskie Problemy Razrabotki Poleznykh
With use of DT and Do, Eqs.[A12] and [A13] then Iskopaemykh, 5, 104-109, Sept.-Oct., 1989.
become for the two layers (i = 1, 2; j ≠ i): 5. Bachrach, R., Nur, A., and Agnon, A., "Liquefaction and
dynamic poroelasticity in soft sediments", J. Geophys.
∂ pi Res., 106(B7), 13,515-13,526 (2001); Correction to the
1 ∂σ ∂ ⎡ ∂ pi ⎤ DT
Ai − − λ − ( p − p j ) = 0 [A16] paper, 108(B3), 2145 (2003).
∂ t M i ∂ t ∂ x ⎢⎣ Ti ∂ x ⎥⎦ hi i 6. Barabanov, V. L., Grinevskiy, A. O., Kissin, I. G., and
Nikolayev, A. V., “Some Effects of Imposed Seismic
12 SPE 103870

Vibrations on a Water-Saturated Medium. Comparison 23. Dusseault, M. B., “Cold Production and Enhanced Oil
with Effects of Distant Earthquakes”, Dokl. USSR Acad. Recovery”, J. Canadian Petrol. Tech., 32 (9), 16-18,
Sci., Earth Sci. Sect., 297, 52-56 (1987). September 1993.
7. Bedford, A., Costley, R. D., and Stern, M., “On the Drag 24. Ellingsen, O., "EOR by Electro-Acoustic Reservoir
and Virtual Mass Coefficients in Biot’s Equations”, J. Stimulation: A New Approach", World Oil, 29-35, Nov.
Acoust. Soc. Amer., 76(6), 1804-1809 (1984). 2002.
8. Beresnev, I. A., and Nikolaevskiy, V. N., “The Effect of 25. Ganiev, R. F., Ukrainskii, L. E., and Frolov, K. V., "Wave
the Nonlinearity Coefficient in the Excitation of Mechanism for the Acceleration of a Liquid Flowing in
Dominant Seismic Frequencies”, Dokl. USSR Acad. Sci., Capillaries and Porous Media", Soviet Phys. Dokl.,
Earth Sci. Sect., 317, 1103-1107 (1991). 34(6), 519-521 (1989).
9. Beresnev, I. A., and Johnson, P. A., “Elastic-wave 26. Ganiev, R. F., and Ukrainskii, L. E., "Monoharmonic
stimulation of oil production: A review of methods and Self-Excited Oscillations Bifurcating from Poiseulle Flow
results”, Geophysics, 59(6), 1000-1017 (1994). in a Compliant Pipe of Circular Cross Section", Fluid
10. Beresnev, I. A., Vigil, R. D., Li, W., Pennington, W. D., Dyn., 26(4), 501-508 (1992).
Turpening, R. M., Iassonov, P. P., and Ewing, R. P., 27. Gao, C.-T., and Deans, H. A., “Pressure Transients and
“Elastic waves push organic fluids from reservoir rock”, Crossflow Caused by Diffusities in Multilayer
Geophys. Res. Lett., 32, L13303 (2005). Reservoirs”, SPE Formation Evaluation, 438-448, June
11. Biot, M. A., “Theory of Propagation of Elastic Waves in a 1988.
Fluid-Saturated Porous Solid. I. Low-Frequency Range”, 28. Graham, D. R., and Higdon, J. J. L., "Oscillatory Flow of
J. Acoust. Soc. Amer., 28(2), 168-178 (1956a). Droplets in Capillary Tubes. Part 1. Straight Tubes", J.
12. Biot, M. A., “Theory of Propagation of Elastic Waves in a Fluid Mech., 425, 31-53 (2000a).
Fluid-Saturated Porous Solid. II. Higher Frequency 29. Graham, D. R., and Higdon, J. J. L., "Oscillatory Flow of
Range”, J. Acoust. Soc. Amer., 28(2), 179-191 (1956b). Droplets in Capillary Tubes. Part 2. Constricted Tubes",
13. Booth, S., and Araujo, M., “Wettability effects on cyclic J. Fluid Mech., 425, 55-77 (2000b).
waterflooding—Experimental sandpack study”, P07, 30. Groenenboom, J., Wong, S.-W., Meling, T., Zschuppe,
presented at 13th Euro. Symp. Improved Oil Recovery, R., and Davidson, B., “Pulsewd water injection during
Budapest, Hungary, Apr. 25-27, 2005. waterflooding”, SPE 84856, presented at Intern.
14. Bourbie, T., Coussy, O., and Zinszner, B., Acoustics of Improved Oil Recovery Conf. Asia Pacific, Kuala
Porous Media, Gulf Publishing, Houston (1987). Lumpur, Malaysia, Oct. 20-21, 2003.
15. Brunner, W. M., and Spetzler, H. A., "Observations of 31. Hilpert, M., Stopper, D., and Jirka, G. H., "Resonance of
Time-Dependent Meniscus Behavior with Implications a Liquid Column in a Capillary Tube", Z. angew. Math.
for Seismic Attenuation in Three-Phase Systems", Phys., 48, 424-438 (1997).
Geophys. Res. Lett., 28(9), 1867-1870 (2002). 32. Hilpert, M., and Miller, C. T., "Experimental
16. Cable, A. S., Dorey, M. P., Goodyear, S. G., and Investigation on the Resonance of a Liquid Column in a
Jayasekera, A. J., "Pulse Injection Technology for IOR", Capillary Tube", J. Colloid Interface Sci., 219, 62-68
Paper IOR-07 for 11th European Symp. Improved Oil (1999).
Recovery, Amsterdam, The Netherlands, June 11-12, 33. Hilpert, M., Jirka, G. H., and Plate, E. J., "Capillary-
2001. Induced Resonance of Oil Blobs in Capillary Tubes and
17. Charlaix, E., Kushnick, A. P., and Stokes, J. P., Porous Media", Geophysics, 65(3), 874-883 (2000).
“Experimental Study of Dynamic Permeability in Porous 34. Hirasaki, G., and Zhang, D. L., “Surface Chemistry of Oil
Media”, Phys. Rev. Lett., 61 (14) 1595-1598 (1988). Recovery from Fractured, Oil-Wet, Carbonate
18. Charlaix, E., and Gayvallet, H., "Dynamics of a Formations”, SPE Journal, 151-162, June 2004.
Harmonically Driven Fluid Interface in a Capillary", J. 35. Huh, C., Wylie, P. L., Jr., Shyeh, J. J., and Bailey, J. R.,
Phys. II France, 2, 2025-2038 (1992). “Method for Improving Oil Recovery by Delivering
19. Cheremisin, N. A., Sonich, V. P., and Efimov, P. A., Vibrational Energy in a Well Fracture”, U. S. Patent
“Influence of Collector’s Irreversible Deformation on 6,814,141 B2, Nov. 9, 2004.
Oil”, IOR-14, presented at 11th Euro. Symp. Improved 36. Iassonov, P. P., and Beresnev, I. A., "A model for
Oil Recovery, Amsterdam, The Netherlands, June 11-12, enhanced fluid percolation in porous media by application
2001. of low-frequency elastic waves", J. Geophys. Res.,
20. Delshad, M., "Trapping of micellar fluids in Berea 108(B3), 2001JB000683 (2003).
sandstone", Ph.D. Thesis, University of Texas at Austin 37. Ivanov, D. A., and Araujo, M. G., "Dynamics of Two-
(1990). Phase Pulsed Flow", SPE 99678, presented at SPE/DOE
21. Dimon, P., Kushnick, A. P., and Stokes, J. P., "Resonance 15th Symp. Improved Oil Recovery, Tulsa, OK, Apr. 22-
of a Liquid-Liquid Interface", J. Phys. France, 49, 777- 26, 2006.
785 (1988). 38. Jin, Y., Ma, J., Jiang, H., Li, X., Wang, Z., and Chang, L.,
22. Dobronravov, O. V., "A new technology of reservoir "Preliminary experimental study of percolation under the
stimulation through exposure to weak seismic waves", condition of vibration", Oil Drilling Prod. Tech., 19, 28-
First Break, 20, 376-382 (2002). 31 (1997).
SPE 103870 13

39. Jin, Y., Jiang, H., and Yin, L., "Microexperimental study 54. Ma, J., Jin, Y., Zhou, S., and Zhai, G., "Experiments on
of the mechanism of vibration production", Oil Drilling the effects of mechanical vibration on core permeability",
Prod. Tech., 19, 32-35 (1997). J. Xi’an Petrol. Inst., 11(5), 8-11, 15 (1997).
40. Kostrov, S. A., Wooden, W. O., and Roberts, P. M., "In 55. Miksis, M. J., “Effects of Contact Line Movement on the
Situ Seismic Shockwaves Stimulate Oil Production", Oil Dissipation of Waves in Partially Saturated Rocks”, J.
& Gas J., Sept. 3, 2001. Geophys. Res., 93 (B6), 6624-6634 (1988).
41. Kouznetsov, O. L., Simkin, E. M., Chilingar, G.V., and 56. Moerig, R., Waite, W. F., Boyd, O. S., Getting, I. C., and
Katz, S. A., "Improved Oil Recovery by Application of Spetzler, H. A., "Seismic Attenuation in a Partially
Vibro-energy to Waterflooded Sandstones", J. Petrol. Saturated, Artificial Crack Due to Restricted Contact Line
Sci. Eng., 19, 191-200 (1998). Motion", Geophys. Res. Lett., 24(24), 3309-3312 (1997).
42. Krylov, A. L., Nikolaevskiy, V. N., and El, G. A., 57. Nikolaevskiy, V. N., “Mechanism and dominant
“Mathematical Model of Linear Ultrasonic Wave frequencies of vibrational enhancement of yield of oil
Generation by Seismic Waves”, Dokl. USSR Acad. Sci., pools”, Dokl. USSR Acad. Sci., Earth Sci. Sect., 307,
Earth Sci. Sect., 318, 1340-1345 (1991). 570-575 (1989).
43. Kurlenya, M. V., and Serdyukov, S. V., "Low-Frequency 58. Nikolaevskiy, V. N., “The Effect of Vibration on the
Resonance of Seismic Luminescence of Rocks in a Low- Ultimate Recovery of Oil from Reservoirs”, The
Energy Vibration-Seismic Field", J. Mining Sci., 35(1), Mechanics of Liquids and Gas, No.5, UDK 532.546:
1-5; "Study of the Formation and Relaxation Processes of 534.2-18 (1992a).
Seismic Luminescence of Rocks in a Low-Energy 59. Nikolaevskiy, V. N., “Rock Vibration and Finite Oil
Vibration-Seismic Field", ibid, 6-11 (1999a). Recovery”, Fluid Dynamics, 27(5), 689-696 (1992b).
44. Kurlenya, M. V., and Serdyukov, S. V., "Nonlinear 60. Nikolaevskiy, V. N., Lopukhov, G. P., Liao, Yizhu, and
Effects in the Radiation and Propagation of Vibro- Economides, M. J., “Residual Oil Reservoir Recovery
Seismic Signals in a Rock Mass", J. Mining Sci., 35(2), With Seismic Vibrations”, SPE Production & Facilities,
105-112; "Reaction of Fluids of an Oil-Producing Stratum 89-94, May 1996.
to Low-Intensity Vibro-Seismic Action", ibid, 113-119 61. Norris, A. N., “Low-Frequency Dispersion and
(1999b). Attenuation in Partially Saturated Rocks”, J. Acoust. Soc.
45. Kuznetsov, O. L., Simkin, E. M., Chilingar, G. V., Amer., 94(1), 359-370 (1993).
Gorfunkel, M. V., and Robertson, Jr., J. O., "Seismic 62. Pan, Y., and Horne, R. N., "Resonant Behavior of
Techniques of Enhanced Oil Recovery: Experimental and Saturated Porous Media", J. Geophys. Res., 105(B5),
Field Results", Energy Sources, 24, 877-889 (2002). 11,021-11,028 (2000).
46. Kuznetsov, O. L., and Simkin, E. M., Transformation 63. Pogosyan, A. B., Simkin, E. M., Stremovskiy, E. V.,
and Interaction of Geophysical Fields in Lithosphere, Surguchev, M. L., and Shnirel’man, E. M., “Separation of
Cambridge Univ. Press, Cambridge, 1991. hydrocarbon fluid and water in an elastic wavefield acting
47. Li, Z., Shao, X., Qiu, Y., and Chen, X., “Methods and on a porous reservoir medium”, Dokl. USSR Acad. Sci.,
effects of cyclic waterflooding in the Southern oilfields of Earth Sci. Sect., 307, 575-577 (1989).
Daqing Placanticline”, SPE Adv. Tech. Series, 5(1), 20- 64. Pope, G. A., and Baviere, M., “Reduction of Capillary
23 (1995). Forces by Surfactant”, in Basic Concepts in Enhanced
48. Li, W., Vigil, R. D., Beresnev, I. A., Iassonov, P., and Oil Recovery Processes, 89-122, Baviere, M., ed.,
Ewing, R., "Vibration-induced mobilization of trapped oil Elsevier, 1991.
ganglia in porous media: Experimental validation of a 65. Reddi, L. N., "Feasibility of In Situ Implementation of
capillary-physics mechanism", J. Colloid Interface Sci., Vibrations to Mobilize NAPL Ganglia", J. Soil
289, 193-199 (2005). Contamin., 3(1), 29-46 (1994).
49. Li, X., Zhong, L., and Pyrak-Nolte, L. J., "Physics of 66. Reddi, L. N., and Challa, S., "Vibratory Mobilization of
Partially Saturated Porous Media: Residual Saturation and Immiscible Liquid Ganglia in Sands", J. Environ. Eng.,
Seismic-Wave Propagation", Annual Rev. Earth Planet. 120(5), 1170-1190 (1994).
Sci., 29, 419-460 (2001). 67. Reddi, L. N., and Wu, H., "Mechanisms Involved in
50. Lo, W.-C., Sposito, G., and Majer, E., “Wave propagation Vibratory Destabilization of NAPL Ganglia in Sands", J.
through elastic porous media containing two immiscible Environ. Eng., 122(12), 1115-1119 (1995).
fluids”, Water Resources Res., 41, W02025 (2005). 68. Roberts, P. M., Venkitaraman, A., and Sharma, M. M.,
51. Lo, W.-C., and Sposito, G., “Immiscible two-phase fluid "Ultrasonic Removal of Organic Deposits and Polymer-
flows in deformable porous media”, Adv. Water Induced Formation Damage", SPE Drilling &
Resources, 25, 1105-1117 (2002). Completions, 15-1, 19 (2000).
52. Lopuchov, G. P., “Vibroseismic Stimulation for 69. Roberts, P. M., Sharma, A., Uddameri, V., Monagle, M.,
Rehabilitation of Waterflooded Reservoirs”, Petroleum Dale, D. E., and Steck, L. K., "Enhanced DNAPL
Geoscience, 5, 259-263 (1999). Transport in a Sand Core during Dynamic Stress
53. Ma, J., Liu, W., Huang, H., Liu, S., Zhang, H., and Stimulation", Environ. Eng. Sci., 18-2, 67-79 (2001).
Zhang, Y., "Increasing the waterflood recovery efficiency 70. Roberts, P. M., Esipov, I. B., and Majer, E. L., "Elastic
of core by mechanical vibration", J. Xi’an Petrol. Inst., Wave Stimulation of Oil Reservoirs: Promising EOR
12(4), 19-21 (1997). Technology?", The Leading Edge, 448-453, May 2003.
14 SPE 103870

71. Roberts, P. M., "Laboratory observations of altered 85. Stirpe, M. T., Guzman, J., Manrique, E., and Alvarado,
porous fluid-flow behavior in Berea sandstone induced by V., “Cyclic water injection simulations for evaluation of
low-frequency dynamic stress stimulation", Acoustic its potential in Lagocinco Field”, SPE 89378, presented at
Phys., 51(7), S140- S148 (2005). SPE/DOE 14th Symp. Improved Oil Recovery, Tulsa,
72. Santos, J. E., Douglas, J., Jr., and Corbero, J., "Static and OK, Apr. 17-21, 2004.
dynamic behavior of a porous solid saturated by a two- 86. Surguchev, L., Koundin, A., Melberg, O., Rolfsvag, T.
phase fluid", J. Acoust. Soc. Amer., 87(4), 1428-1438 A., and Menard, W. P., “Cyclic water injection: improved
(1990a). oil recovery at zero cost”, Petrol. Geosci., 8, 89-95
73. Santos, J. E., Douglas, J., Jr., Corbero, J., and Lovera, O. (2002).
M., "A model for wave propagation in a porous medium 87. Tuncay, K., and Corapcioglu, M. Y., "Body Waves in
saturated by a two-phase fluid", J. Acoust. Soc. Amer., Poroelastic Media Saturated by Two Immiscible Fluids",
87(4), 1439-1448 (1990b). J. Geophys. Res., 101(B11), 25,149-25,159 (1996).
74. Santos, J. E., Ravazzoli, C. L., Gauzellino, P. M., 88. Tuncay, K., and Corapcioglu, M. Y., "Wave Propagation
Carcione, J. M., and Cavallini, F., "Simulation of waves in Poroelastic Media Saturated by Two Fluids", J. Appl.
in poro-viscoelastic rocks saturated by immiscible fluids. Mech., 64, 313-320 (1997).
Numerical evidence of a second slow wave", J. Comput. 89. Wang, J., Dusseault, M. B., Spanos, T., and Davidson, B.,
Acoust., 12(1), 1-21 (2004). "Fluid Enhancement Under Liquid Pressure Pulsing at
75. Schutt, H., Kohler, J., Boyd, O. S., and Spetzler, H. A., Low Frequency", Paper 1998.016 for 7th UNITAR, 151-
"Seismic Attenuation in Partially Saturated Dime-Shaped 159, 1998.
Cracks", Pure Appl. Geophys., 157, 435-448 (2000). 90. Westermark, R. V., Brett, J. F., and Maloney, D. R.,
76. Seethepalli, A., Adibhatla, B., and Mohanty, K. K., "Using Downhole Vibration Stimulation for Enhanced Oil
“Physicochemical Interactions during Surfactant Flooding Recovery", World Oil, 57-66, Oct. 2001.
of Fractured Carbonate Reservoirs”, SPE Journal, 411- 91. Westermark, R. V., and Brett, J. F., "Enhanced Oil
418, Dec. 2004. Recovery Using Downhole Vibration Stimulation, Osage
77. Siminov, V. F., Serdyukhov, S. V., Cherednikov, E. N., County, Oklahoma", SPE 75254, presented at SPE/DOE
Sibirev, A. P., Nozhin, V. M., Lepikhin, A. G., and 13th Symp. Improved Oil Recovery, Tulsa, OK, Apr. 13-
Kaniskii, N. A., “Results of Experimental Oilfield Study 17, 2002.
on Enhancing Oil Recovery by Vibroseismic Method”, 92. White, J. E., "Computed Seismic Speeds and Attenuation
Oilfield Development and Production, 48-52 (1996). in Rocks with Partial Gas Saturation", Geophysics, 40(2),
78. Simkin, E. M., and Surguchev, M. L., “Advanced 224-232 (1975).
Vibroseismic Technique for Water Flooded Reservoir 93. White, J. E., Mikhaylova, N. G., and Lyakhovitskiy, F.
Stimulation, Mechanism and Field Test Results”, 6th M., “Low-Frequency Seismic Waves in Fluid-Saturated
European IOR Symp., Stavanger, Norway, May 21-23, Layered Rocks”, Izv. Earth Phys., 10, 44-52 (1975).
1991. 94. Xue, Z., and Yang, W., "Unconventional Stimulation
79. Simkin, E. M., Surguchev, M. L., Akhapkin, M. Yu., Technologies – State of the Art and Considerations for
Pogosyan, A. B., and Stupochenko, V. E., “Effect of Further Development", J. Xi’an Petrol. Inst., 11(1), 15-
Elastic Fluctuations on the Capillary Water Impregnation 18 (1996).
of Oil-Saturated Porous Media”, Soviet Phys. Dokl., 95. Yuan, Q., Ji, B., and Yang, J., “Mechanism of Cyclic
36(4) 295-6, 1991. Waterflooding in Vertically Heterogeneous Reservoirs”,
80. Smirnova, M. N., “Effect of Earthquakes on the Oil Yield SPE Adv. Tech. Series, 5(1), 24-27 (1995).
of the Gudermes Field (Northeastern Caucasus)”, Izv. 96. Zhang, L. H., Ho, P., Li, Y., and He, S., “Low-Frequency
Earth Physics, 12, 71-76 (1968). Vibration Recovery Enhancement Process Simulation”,
81. Snarsky, A. N., “Determination of the Effect of an SPE 51914, presented at SPE Reservoir Simul. Symp.,
Infrasound Field on the Flow Rate of Oil Using an Houston, TX, Feb. 14-17, 1999.
Elementary Reservoir Model”, Trans. Higher 97. Zhu, T., Xutao, H., and Vajjha, P., “Downhole harmonic
Educational Inst., Oil and Gas (1), UDC 622.276.03, vibration oil-displacement system: A new IOR tool”, SPE
1982. 94001, presented at SPE Western Regional Meet., Irvin,
82. Spanos, T., Davidson, B., Dusseault, M. B., and Samaroo, CA, Mar. 30-Apr. 1, 2005.
M., "Pressure Pulsing at the Reservoir Scale: A New IOR
Approach", Paper 99-11 for CSPG and Canadian
Petrol. Soc. Joint Convention, Calgary, Alberta, June
14-18, 1999.
83. Spanos, T., Davidson, B., Dusseault, M., Shand, D., and
Samaroo, M., “Pressure pulsing at the reservoir scale: A
new IOR approach”, J. Canadian Petrol. Tech., 42, 16-
28 (2003).
84. Spanos, T. J. T., The Thermophysics of Porous Media,
Chapman & Hall/CRC Monographs and Surveys in Pure
and Applied Mathematics 126, Chapman & Hall/CRC,
2002.
SPE 103870 15

r = 400 m icron

1.2 1
Berea sandstone-
Norm. Residual Oil Saturation

1 Delshad 1990
Model 0.8

0.8
0.6

So / Sori
0.6

0.4
0.4 200 Hz
100 Hz
0.2 0.2
50 Hz
20 Hz
0
0
1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00
1 10 100 1000
Capillary Number
Displacement (micron)

Figure 1: Capillary number correlation. Figure 4: Residual oil displacement efficiency in terms of
rock displacement (ε) and frequency (ω), for r = 400 μm.
r = 100 micron
1

0.8
σ
0.6 200 Hz P1 K1 SO1
So / Sori

h1
100 Hz
0.4 50 Hz
P2 K2 SO2
h2
0.2

0
1 10 100 1000
Displacement (micron )

Figure 2: Residual oil displacement efficiency in terms of Figure 5: Schematics of the simplified two-layer reservoir
rock displacement (ε) and frequency (ω), for r = 100 μm. model with application of rock normal stress oscillation, σ.

r = 200 micron 0.08

1

0.8 0.06
(P1-P2)/σa

0.6
So / Sori

0.04

0.4
200 Hz k1=1; k2=0.01
0.02
100 Hz
0.2 k1=1; k2=0.05
50 Hz
20 Hz k1=5; k2=0.25
0 0
1 10 100 1000 1 10 100 1000
Displacement (micron) Frequency (Hz)

Figure 3: Residual oil displacement efficiency in terms of Figure 6: Modulus of pressure differential between two layers
rock displacement (ε) and frequency (ω), for r = 200 μm. (scaled with σa) in terms of frequency (ω) and different
combinations of layer permeabilities (for So1 = 0.35; So2 =
0.60).
16 SPE 103870

0.15 0.1
k1=1; k2=0.01
k1=1; k2=0.05
k1=5; k2=0.25
0.1 0.01

So rate/day
(P2-P1)/σa

0.05 0.001
So1=0.35; So2=0.60
So1=0.40; So2=0.55
So1=0.45; So2=0.50
0 0.0001
1 10 100 1000 1 10 100 1000
Frequency (Hz) Frequency (Hz)

Figure 7: Modulus of pressure differential between two layers Figure 10: Modulus of instantaneous oil transfer rate between
(scaled with σa) in terms of frequency (ω) and different two layers (scaled with σa) in terms of frequency (ω) and
combinations of layer permeabilities (for So1 = 0.30; So2 = different combinations of oil saturations.
0.80).

0.1 0.1

0.08

0.01
So rate/day

0.06
(P1-P2)/σa

0.04
0.001
k1=1; k2=0.01 So1=0.35; So2=0.60
0.02 k1=1; k2=0.05 So1=0.40; So2=0.55
k1=5; k2=0.25 So1=0.45; So2=0.50
0 0.0001
1 10 100 1000 1 10 100 1000
Frequenct (Hz) Frequency (Hz)

Figure 8: Modulus of pressure differential between two layers Figure 11: Modulus of instantaneous oil transfer rate between
(scaled with σa) in terms of frequency (ω) and different two layers (scaled with σa) in terms of frequency (ω) and
combinations of layer permeabilities (Figure 6 case, with Ko = different combinations of oil saturations (Figure 10 case, with
0.75 x 105 psi instead of 1.0 x 105 psi). Ko = 0.75 x 105 psi instead of 1.0 x 105 psi).

0.12 0.1

0.1

0.08 0.01
So rate/day
(P1-P2)/σa

0.06

0.04 0.001
k1=1; k2=0.01 So1=0.35; So2=0.60

0.02 k1=1; k2=0.05 So1=0.40; So2=0.55

k1=5; k2=0.25 So1=0.45; So2=0.50
0 0.0001
1 10 100 1000 1 10 100 1000
Frequency (Hz) Frequency (Hz)

Figure 9: Modulus of pressure differential between two layers Figure 12: Modulus of instantaneous oil transfer rate between
(scaled with σa) in terms of frequency (ω) and different two layers (scaled with σa) in terms of frequency (ω) and
combinations of layer permeabilities (Figure 6 case, with M1 = different combinations of oil saturations (Figure 10 case, with
5.0 x 105 psi instead of 6.09 x 105 psi). M1 = 5.0 x 105 psi instead of 6.09 x 105 psi).