Professional Documents
Culture Documents
Cyclic Water Injection, Improved Oil Recovery at Zero Cost
Cyclic Water Injection, Improved Oil Recovery at Zero Cost
90 L. Surguchev et al.
zone of higher permeability. Water flow capacity from the low period, between the two layers is (Sharbatova & Surguchev
permeable zone during the depressurizing half-cycle is limited 1988; Gorbunov et al. 1977):
by (a) the relative permeability to water at low water saturation;
(b) the capillary pressure in water-wet rock.
Acceleration of the imbibition and fluid redistribution (cross-
flow between zones of high and low oil saturation) is a goal of
B
?p?= Gq(),
2
?q?= SD 2
3/2
BDkGq()
&k H , &h ,
time of the pressure and it accounts for the compressibility hi
Ki=ki/k̄, k̄= i H= i =x/l, =2/T
effects. It assumes that the total thickness is small compared to i i
the total length of the reservoir and that the porosities and
compressibilities of the layers, viscosities of the fluids, are
equal. The absolute values of the pressures are obtained by multi-
The considered boundary conditions are plication of the dimensionless values with the average value at
the inlet, p*, and the rates (velocities) are obtained by multipli-
cation of the dimensionless values (Gorbunov et al. 1977) with
p(x,t)?x=0=1+Bsint, p(x,t)?x=l=0 (1)
dp(x,t)
dx
U x=0=1+Bsint,
dp(x,t)
dx
U =1
x=l
(2) u*=
Hp*k̄
l 2µ
[3] (4)
For the two types of boundary conditions (equations 1 and The suggested model accounts for redistribution of the oil
2) the dimensionless cross-flows, averaged over the time and water phases in the two-layer system if there is an
Downloaded from http://pg.lyellcollection.org/ at University of Otago on March 14, 2015
additional trapping mechanism like non-linearity, in particular, with the increasing pulse period. From the analytical evaluation
hysteresis, so that the phase mobility depends on the saturation, and field experience of cyclic waterflooding, the optimal value
saturation history and, eventually, on the flow direction. of is in the range of 2 to 8 (Surguchev 1985; Sharbatova &
Surguchev 1988; Gorbunov et al. 1977). The influence of a
number of other parameters – porosity, compressibility,
SIMULATION MODEL FOR CYCLIC INJECTION viscosity and average permeability – is reflected in equation (3),
A 3D model of the Lower Tilje/Åre formations for the I together with the pulse frequency as a combination, dimension-
segment built by the ‘Marginal Facies Team’ (MFT) in 1998 was less parameter .
used in the simulation study. The MFT consisted of members The time period of the pressuring/depressurizing half-cycle
from both Statoil and Conoco. Both one and two pairs of for Lower Tilje/Åre formations is estimated at 5–15 days. As
injection and production wells including existing wells were discussed above, the half-cycle injection time depends mainly
evaluated in simulations (Fig. 1). The simulation model on the distance from the injection well to the oil displacement
was reviewed and adjusted in order to account for pressure front, oil and rock compressibility, permeability values and its
pulsation and capillary pumping effects. variation. Figure 3 shows the effect of these parameters on the
In order to design cyclic injection strategies for the Lower half-cycle time estimation. The reservoir conditions with lower
Tilje/Åre formations evaluation of reservoir parameters and piezoconductivity (=k/µC) and higher compressibility will
calibration of optimal cyclic conditions were done using require longer pressuring/depressurizing half-cycles.
analytical methods. The Lower Tilje/Åre reservoirs are highly The reservoir permeability in the Lower Tilje/Åre forma-
heterogeneous, with large permeability contrasts vertically and tions varies in a quite wide range from a few milli Darcies to
horizontally (Table 1). several Darcies. Under cyclic injection, reservoir pressure can
drop below the bubble point pressure, resulting in the release of
solution gas and increased fluid compressibility. Under such
CYCLIC INJECTIONS FOR LOWER TILJE/ÅRE circumstances the half-cycle time period of pulse injection can
RESERVOIRS vary from some days to some months. The objective of the
In order to evaluate cyclic parameters for the Lower Tilje/Åre numerical simulations is to identify more specifically the opti-
formations the reservoir characteristics were averaged for the mal cyclic conditions for the Lower Tilje/Åre formations at
two-layer system consisting of high permeability and low Heidrun, and to evaluate the potential of cyclic injection in
permeability zones. Reservoir properties used in the analytical terms of additional oil and reduced water production.
evaluation of cyclic process are given in Table 2. Different injection pulses and well placement scenarios are
Figure 2 shows the maximum possible cross-flow between simulated in order to evaluate the cyclic effect and optimize
the high and low permeability zones vs. distance from the injection parameters.
injection well. Assuming complete retention of the water in
the low permeability zones and no permeability restriction the CYCLIC INJECTION WITH PRESSURE PULSES
maximum possible cross-flow is estimated at 1–5 m2 per day
per 1 m width of the oil–water interface area. The maxi- The I segment of the Lower Tilje/Åre simulation model
mum cross-flow value is reduced with time and water front includes one existing production well A-P. The injection well
advancement in the reservoir. A-I is planned to be drilled (Fig. 2). In the base case simulation
The key parameter of cyclic injection is the frequency of scenario of 10 years the production/injection of the wells is
changing cycles (). The cross-flow has a tendency to decrease constrained by bottom hole pressure (BHP = 160 bar) in the
production well and by rate (Qinj = 750 Sm3 per day) in the
injection well. The simulation can be considered as giving a
Table 2. Averaged parameters of Lower Tilje/Åre used in analytical evaluation of cyclic conservative estimate of cyclic injection potential since well
injection production is controlled by BHP and not by WHP. The
Reservoir characteristic Value reduction of production water cut anticipated by cyclic injection
will contribute to better lift and higher oil rate production. In
Oil compressibility (104 bar1) 1.257 these simulations BHP control of production wells was used in
Porosity () (fraction) 0.3 order to maintain the voidage displacement conditions and
Oil viscosity (cp) 2.5 separate the effect of possible increase of water volume
High permeability (k1) (D) 0.6 injection on oil production increase.
Averaged high permeability layer (K1) 0.968 The cumulative oil production after a 10-year period under
Low permeability (k2) (D) 0.020
Averaged low permeability layer (K2) 0.032
base case waterflooding is 1 442 927 Sm3, or 11.3% of
Heterogeneity parameter (V ) 0.031 STOOIP. The simulated pulse injection scenarios can be
Piezoconductivity (= k/µC) (sm2 s1) 5079 divided into four groups by ‘injection/no injection’ time ratios:
2:1, 1:1, 1:2 and 1:3 (Table 3).
Downloaded from http://pg.lyellcollection.org/ at University of Otago on March 14, 2015
92 L. Surguchev et al.
In all simulation runs the total cumulative water injection The cumulative oil production after the 10-year period under
over the 10-year period was kept equal in all scenarios. The base case waterflooding is 2 532 627 Sm3, or 19.9% of
simulation results are shown in Figures 4–6 in comparison with STOOIP. The cumulative oil production of A-P is 839 201 Sm3
base case waterflooding without cycling (w/o cyclic). Cyclic and of A-P2 is 1 693 426 Sm3.
injection scenarios with shorter but more intensive pulsing The simulated cyclic injection scenarios can be divided into
periods (1:2, 1:3) allow oil production to increase by 5.6% over seven groups by ‘injection/no-injection’ time ratios: 2:1, 1:1,
a 10-year period. This translates into 81 000 Sm3 of extra oil 1:2, 1:3, and with injection cycles shifted in time 1:1, 1:2, 1:3
produced by A-P. (Table 4). The simulation results are shown in Figures 7–9. The
best cyclic injection scenarios are those with the shortest cycles
and the most intensive pulsing (1:2, 1:3), which increase oil
ALTERNATING WATERFLOOD PATTERNS AND production by 5.2% over a 10-year production period. The
PRESSURE PULSES incremental oil production for two production wells is about
130 000 Sm3.
The simulated scenarios included the production well A-P and The scenarios with injection cycles shifted in time allow for
the injection well A-I, new production well A-P2 and new a varying waterflood pattern that give higher oil production
injection well A-I2. than simultaneous cyclic injection schemes (Figs 7, 8).
Different injection scenarios and sensitivities were simulated:
injection rates varied from 500 to 3000 m3 per day; EVALUATION OF CYCLIC INJECTION
different cyclic/shut-in periods;
alternating waterflood patterns; The evaluation of cyclic injection in the Lower Tilje/Åre
production well BHP varied from 120 to 200 bar. formations has shown that its main benefits compared with
Downloaded from http://pg.lyellcollection.org/ at University of Otago on March 14, 2015
traditional waterflooding are accelerated oil production, in- years. The oil recovery factor is improved by 0.6% of STOOIP
creased oil reserves/improved sweep efficiency, and reduced after 5 years and 0.7% of STOOIP after ten years of cyclic
water cut. An additional 77 000 Sm3 of oil, or 7.9% increase in injection.
produced reserves, can be achieved for the A-P well after 5 In the cyclic injection scenario with two injection wells and
years of cyclic water injection in A-I injection well in the I two production wells in the I segment the incremental oil
element of the Heidrun Field (Table 5, Fig. 9). After 10 years production after 5 years is 53 000 Sm3 (+2.6%) and after 10
the incremental production from the A-P well amounts to years is 131 000 Sm3 (+5.2%). The oil recovery factor is
81 000 Sm3. The respective reduction in cumulative water improved by 0.4% of STOOIP after 5 years and 1.0% of
production is 64 000 Sm3 after 5 years and 96 000 Sm3 after 10 STOOIP after ten years of cyclic injection.
Downloaded from http://pg.lyellcollection.org/ at University of Otago on March 14, 2015
94 L. Surguchev et al.
We are indebted to the Statoil Heidrun division for supporting this : porosity
work and to Kjell Christoffersen and Kjell Erik Wennberg for useful : dimensionless distance
suggestions. : angular frequency of the pulses at the injection boundary
Subscripts
LIST OF SYMBOLS
i: initial, layer index
B: dimensionless amplitude
o: oil
C: compressibility
g: acceleration of gravity
h: thickness
H: total reservoir thickness REFERENCES
k: permeability Gorbunov, A.T., Surguchev, M. & Tsinkova, O.E. 1977. Cyclic Waterflooding of
KX: absolute permeability along layers Oil Reservoirs. VNIIOENG publication, Moscow (in Russian).
kr: relative permeability Owens, W.W. & Archer, D.L. 1996. Waterflooding Pressure Pulsing for
Fractured Reservoirs. Journal of Petroleum Technology, 745–752.
l: reservoir length
Peng, C.P. & Yanosik, J.L. 1988. Pressure Pulsing Waterflooding in Dual
L: front zone length Porosity Naturally Fractured Reservoirs. Paper presented at the SPE
p: dimensionless pressure International Petroleum Engineering Meeting, China, November 1–4,
q: dimensionless rate 389–400.
t: time Sharbatova, I.N. & Surguchev, M. 1988. Cyclic influence on heterogeneous Oil
T: pulsing injection period Reservoirs. Nedra Publishing House, Moscow (in Russian).
u: Darcy velocity Surguchev, M. 1985. Methods of Secondary and Tertiary Oil Recovery. Nedra
Publishing House, Moscow (in Russian).
x: horizontal (along-dip) co-ordinate
Surguchev, L., Korbøl, R., Haugen, S. & Krakstad, O.S. 1992. Screening of
µ: phase viscosity WAG injection strategies for heterogeneous reservoirs. Paper presented at
: density, also dimensionless frequency the SPE European Petroleum Conference, 16–18 November, Cannes,
: equilibration time constant France.