SPHDOE 35396

Analysis of Hydrofracture Geometry and Matrix/Fracture Interactions During

Steam Injection
A. R. Kovscek, Lawrence Berkeley National Laboratory, R. M. Johnston, CalResources LLC, and T. W. Patzek,
University of California and Lawrence Berkeley National Laboratory, SPE Members
Cqynght 19%5, S,wzcty O( Pelmkum Engmecrs, Inc
?hts paper wm prepared I,Ir prescnlamm m the lY(M SPWOOE Temh Symfnwium m! lmfmvetf
011 Recovery held In Tulsa, OK, 1 I -24 AfwIl 1996
This pqxr WM sclccmd for presentation hy (he SPE Program Commmce following rcwcw of
mformalmn cumainecf In an iihslracl suhmmed by [hc author(s) Contcnls of the pafur as
fmscnmd have mn been mvicwcd by the Sociely of Peuoleum Engineers and arc suhlcc! m
cmrcctum(s) hy [he author(s). The malcrt id, M presented. dot! no{ ncccsswtly rcflccl any
posmon of [he SOch?cy of Pemoleum Engmows or iu mcmfmrs Papers prescnud at SPE meelmgs
are sub)ecl m puhlicalmn rewew hy Wtmrlal Cmmniuecs of fhe SIxIeIy of Petroleum
Engntms. Pcnms.mon u> COPY IS resuuxd m an ahs!racl of nol more than 3(X) wmif.i Illwwrwons
may nm he copf The ahsmacl should comam c<msp:cu(ms acknowkdgmem 01 where and hy
whom [he ~apa was prcscntml WrHc Lthrartan, SPE P O Box 8333836, Richardson, TX 7S083-
M13fI USA, ft. 0-214-952-9435
Abstract
This is the second part of our analysis of a steam drive pilot in
the South Belridge Diatomite, Kern County, California. In
this pilot, steam is injected through two noncommunicating,
vertical hydrofractures (IN2U and IN2L) that nearly span the
entire 1000 ft tall diatomite column. In the first part, we
summarized pilot results for the initial 1200 days of steam
injection and examined steam convection and heating resulting
from injection into the lower hydrofracture.
Here we update the pilot response from 10/90 to 11/95.
We conclude that the cumulative oiI production in the far
(543N) and close (543P) producer was 106,000 and 55,000
BO, respectively, and the incremental oil production above
primary was 60,000 and 37,000 BO, respectively. This
translates to the total oil recovery of 9% OOIP and the
incremental oil recovery over primary of 6% OOIP after 5
years of steam injection. From our simulations of steam
injection in IN2U and IN2L, we estimate that 246,000 BS
CWE was injected to the west towards 543N and 102,000 BS
was injected to the east toward 543P. The cumulative oil
steam ratio (COSR) for the pilot was, therefore, 0.45, and the
incremental COSR was 0.28.
We also apply a high resolution numerical model
developed in the first paper to interpret the results of steam
injection through the upper hydrofmcture of the pilot. Results
of this analysis indicate that the upper injection hydrofiacture
was highly dynamic and asymmetrical while undergoing steam
injection. Steam flowed preferentially into the northern wing
of the hydrofracture which reached a final winglength of 250 ft.
To the south, hydrofracture winglength reached roughly 75 ft
and diatomite heating was weak. A dramatic temperature
response above the perforated interval of the upper injection
well suggests that a horizontal fracture, or network of natural
fractures, opened within the formation and linked to the
injection hydrofracture. Our analysis indicates that roughly 43
percent of the total injected steam migrated above the
perforations of the injection well, but remained within the
Diatomite, and flowed rapidly away from the fracture face b
to a large increase of hydraulic diffusivity of the formation.
Introduction
Thermal oil recovery in shallow, low permeability, and
hydraulically fractured reservoirs appears to be substantially
more efficient than conventional waterflooding. Thus, steam
drive for such reservoirs is under investigation as an alternative
secondary oil recovery process. To help assess the feasibility
of steam drive for oil recovery in diatomaceous oil-hearing
rocks, Shell Western E&P conducted the Phase I and II pilots
[1] in Section 29 of the South Belridge Diatomite, Kern
County, California.
This is the second paper of a two part series presenting
an interpretation of the Phase H pilot. In the pilot, two
separate hydrofractured injectors that are denoted IN2U and
IN2L, where the U and L refer to a single perforation across
the upper and lower portion of the Diatomite, are used to
deliver steam to the formation. In our first paper [2] (referred to
as Part 1), the pilot design, data collection program, and
geology were described along with a computationally simple
model for interpreting the pilot results. Part I also validated the
model by interpreting the results of steam injection through
IN2L.
519
2 ANALYSISOF HYDROFRACTUREGEOMETRY
SPE 35396
AND MATRIWFRACTURF INFRACTIONS DURING STFAM INJFCTION
Here, we concentrate on the results of steam injection
through IN2U which has the dckd complication of a
horizontal fracture extension in addition to an asymmetric
shape of the hydrofkacture. Finally, we compare the inferred
hydrofracture shapes deduced from the earlier microseismic
imaging study with the hydrofracture shapes from history
matching steam injection in IN2U and IN2L. The latter
comparison allows us to judge whether microseismic events
are indicative of heat delivery to diatomite.
First, we review the Phase II pilot design and update
pilot results through the end of 1995. Next, we present a
history match of the Phase 11 response to steam injection
through IN2U. Thirdly, the sequence of hydrofracture shapes
obtained for IN2U and IN2L are compared to the hydrofracture
shapes indicated by passive microseismic imaging of the
hydrofractunng process. Finally, the implications of relative
changes in hydraulic diffusivity, on the productivity of wells
are discussed.
Phase II Steam Drive Pilot
Figure 1 presents a plan view of the Phase II pilot area
including injection, production, and monitoring wells. The
two injectors, IN2U and IN2L, located in the center of the
figure are perforated from 1110 to 1460 ft and 1560 to 1910 ft,
respectively. The dual injectors allow separate control of steam
injection rates and pressures in the upper and lower portion of
the Diatomite. An earlier passive microseismic imaging study
[3] of the hydrofractures in these two wells indicated that while
the hydrofracture in IN2L appeared to be symmetric, that in
IN2U was highly asymmetric. The azimuth of both fractures
was found to be N21 E ~ 4. The dark diagonal line passing
through IN2U and IN2L gives this azimuth direction
graphically.
The close producer, 543P, is 40 ft to the east of the
injection wells and the far producer, 543N, is roughly 130 ft to
the west. Microseismic monitoring in 543P [4] indicated that
the hydrofracture in the well was nearly vertical and divided
into two separate zones corresponding to the most permeable
diatomite layers. Also, the upper most fractured zone extended
60 ft above the pforated zone. The azimuth of the
hydrofracture in 543P is reported as N25E * 5. Well 543N is
an old, full interval prcducer drilled and completed in October
1979. Again, the dark diagonal lines passing through the
production wells give the general direction of the
hydrofractures. The remaining wells, LO 11 to L015, MO] and
M02, are observation wells that measure the formation
temperature response and oil displacement.
Injection in IN2U began on October 10, 1990 and IN2L
followed on October 24, 1990 and steam injection continues as
of the writing of this paper. Daily averages of the minute by
minute measurements of steam injection rate and pressure am
saved for each injection well. Also, the production response is
measured daily and temperature surveys in the observation
wells are run, on average, every 30 days between 800 and 2tXKl
ft at IO ft intervals.
Phase II Response. Significant heating of the formation
and oil production response occurred as a result of steam
injection. Increase in the size of the hydrofracture in IN2U
during steam injection resulted in a roughly 6-fold increase of
injectivity between 100 and 900 days of steam injection, as
shown in Fig. 2. Tip extensions in IN2L also manifested
themselves as substantial increases in steam injectivity [2]. III
Fig. 2, the second derivative of cumulative steam injection in
IN2U with respect to the square root of time increased
constantly at about 600 bbllday between 225 and 900 days. At
slightly more than 900 days, there is a sharp drop in
injectivity due to a near step-change decrmse in injection
pressure; however, the slope of the cumulative injection
quickly returns to roughly 600 bbl/day even though injectivity
decreased as a result of the decrease in injection pressure. The
reduction of injection pressure is also quite evident in the
cumulative injected steam history as a discontinuity at about
900 days. We attribute this last sharp decrease of injectivity to
the shutdown of the horizontal part of the injection fracture.
Both producers, 543P and 543N, show a substantial oil
response to steam injection. The near producer, 543P (Fig, 3),
responded to steam injection within the first 100 days, Its
cumulative oil production roughly tripled compared with that
calculated for a limited-interval primary producer in the M-
cycle that was pressure-depleted by 543N (Table 1). A strong
linkage with the IN2L hydrofracture resulted in steam
breakthrough at the producer and wellhead temperatures
approaching 320 F. To mitigate the steam breakthrough, a
packer was set in 543P that isolated the lower diatomite layers
where steam breakthrough had occurred from the remaining
productive interval of the well. As Fig. 3 shows, wellhead
temperature then decreased to 120 F and well productivity
improved. Even though the perforated interval in 543P spans
only 350 ft, its productivity has been as high 1300 BO/daysn,
In Section 29, average productivity of a full-interval,
undepleted prcducer on primary is 2300 BO/daysn. We
estimate that as of 11/95, 543P produced 37,000 BO above
primary.
The far producer, 543N, whose fractures span 850 ft of
the diatomite column, saw oil production increase from 40
BOPD at the beginning of the pilot to 150 BOPD. llre
extended production history of 543N (Fig. 4), shows some
interesting effects of steam injection on diatomite. Time mm
corresponds to November 1979 when this well was brought on
production. Steam injection begins some ten years later (60
dayslfl) and 543N begins to show oil production response to
steam at 680 days of steam injection (66 days] n). During the
initial period of oil production response, between 66 and 70
days~~, 543N appears to be plugging because the second
520
SPE 35396 A. R. KOVSCEK, R. M. JOHNSTON, AND T. W. PATZEK 3
derivative of production with respect to the square root of time
is -2500 bbl/day. That is, the oil production rate appears to k
beaded toward zero. After 70 days2, the curvature of the
cumulative oil production response changes, and the second
derivative of production with respect to time becomes +1300
bbl/day. This change in sign of the slope indicates that well
productivity has stopped decreasing and has actually begun to
improve Note also, that the improvement in productivity is
accompanied by a slow increase in wellhead temperature from
120 F to approximately 135 F.
Finally, a simple analytical match of oil production in
543N shows (diamonds in Fig. 4) that pressure interference
does become visible after 60 days *D, as the cumulative oil
versus square root of time curve begins to bend down. Thus,
the incremental oil response in 543N is estimated to be 60,000
BO as of 11/95.
Bottom Line. Between 10/1/90 and 11/15/95, 175,000 and
214,000 BS CWE were injected in IN2U and IN2L,
respectively. In the same time period, the cumulative oil
production in 543N and 543P was 106,000 and 55,000 BO,
respectively, and the incremental oil production above primary
was estimated to be 60,000 and 37,000 BO, respectively.
From our simulations of steam injection in IN2U and IN2L,
we estimate that 246,000 BS CWE were injected to the west
towards 543N and 102,000 BS were injected to the east toward
543P (Remember that there is no producer to the east of IN2U
and the pilot geometry causes asymmetrical steam flow.).
Therefore, the cumulative oil steam ratio (COSR) for the pilot
was 0.45, and the incremental COSR was 0,28. The bulk rock
volume of the pilot is calculated as 9,300,000 bbl accounting
for the asymmetrical pilot geometry. We estimate the pore atKI
oil volumes 10 be 4,400,000 and 1,410,000 bbl, respectively,
to the west and 504,000 and 162,000 bbl, respectively, to the
east of the injectors. Hence, the total oil recovery oil rvcovety
in the pilot was 9?Z0of OOIP and the incremental oil recovery
was 690 of OOIP after 5 years of steam injection.
Model Description
Our computational approach is to lump the parameters which
describe the first-order physics of steam movement in
diatomite into a single parameter, termed the hydraulic
diffusivity, thereby simplifying history matching. This
formulation leads to a second order, partial differential
equation, similar in form to a transient diffusion equation, that
describes the pressure profile in the steam-wcupied zone of
each layer of the Diatomite. The method of Marx ml
Langenheim [5] is then used to locate the steam front in mh
layer. Finally, the transient heat conduction equation is solved
for the temperature profile in the oil zone downstream of the
steam front. The hydraulic and thermal diffusivities as well as
hydrofiacture shape are then iterated to match the areal
temperature response in each layer of the Diatomite in the
Phase H pilot and the overall, cumulative steam injection.
Our approach is computationally efficient and leads to
multiple realizations of steamdrive dynamics in the Diatomite.
More importantly, this modeling effort provides the precursory
knowledge of fracture size and shape, as well as the relative
changes in formation permeability needed to simulate
steamdrive in the Diatomite with fully compositional,
multidimensional reservoir simulators.
The pressure diffusion equation is derived by writing
the usual continuum mass and energy balance equations for a
black-oil model reservoir simulation model and expanding
spatial derivatives. Neglecting high order derivative terms,
gravity, capillarity, and assuming saturated steam conditions
[2] yields
32p 1 +
&2 =--~ . . . . . . . . . . . . . . .. . . . . . . ...(l)
where the ratio of the total mobility of all phases, kT, upon
the total system compressibility, v, is the hydraulic diffusivity
Cl=lz
.. ....... . . ... . .
v
(2)
Here p is pressure, x is distance from the injection plane, and t
is time. Because the South Belridge Diatomite is so finely
layered [6] and hydrofracture length is generally greater than the
height of a layer, a one-dimensional equation suffices [2].
The position of the steam front, Xf, within each layer of
the Diatomite is obtained by performing a bulk heat balance in
the horizontal direction and assuming that layer properties am
uniform and constant, After neglecting interlayer heat transfer
we obtain
Xf =
\
Qi ~t
. . .. . . . .. . . . . .. .
(3)
,, MT&AT1
where Qi is the rate of heat injection, MT is the total
volumetric specific heat, Ac is the layer vertical cross sectional
area, and ATI (=T1 - To) is the saturated steam temperature at
the fracture face minus the initial layer temperature.
Over the portions of the formation not swept by steam
heat transport occurs by thermal conduction. Consistent with
Eq. (l), one-dimensional heat conduction is described by
d2T 1 c?T
.
p= K& if-----------
(4)
where K (z~T) is the thermal diffusivity of the formation
521
4 ANALYSIS OF HYDROFRACTURE GEOMETRY SPE 35396
AND MATRI)VFRACTURF INFRACTIONS DURING STEAM INJFCTION
while K is the thermal conductivity.
History Match
To match cumulative steam injection into IN2U and the
temperature response at all monitoring wells in the Phase II
pilot, we follow the history matching procedure as outlined in
Part I and ref. [7]. The diatomite interval between 800 ad
1500 ft in depth is divided into 8 isolated layers of varying
thickness along geologic boundaries. The nodal spacing in the
horizontal or x-direction is 0.66 ft to the west of IN2U while
to the east it is 0.50 ft. Figure 5 displays the porosity log
from Well LO 15, and for comparison, the porosities for each
layer used in the simulations to follow as a solid dark line.
The simulation volume spans the diatomite cycles C through I
as denoted by Schwartz [6]. The light horizontal dashcxi lines
on Fig. 5 indicate the tops of these cycles.
Initial and boundary conditions for IN2U are identical to
the interpretation of the IN2L response. For instance, the
initial temperature is given by Eq. (29) of Part I and a 3.3
psi/ft pressure gradient in the horizontal direction is assigned
across the pilot area. A pressure gradient across the pilot area
exists due to 10 years of pressure depletion caused by
production from 543N. The injection pressure for simulation
is fixed by the injection pressure history for the pilot between
October 1989 and January 1994, as shown in Fig. 6.
Consistent with the analysis of IN2L, we make a capillary-
pressure correction to the downhole injection pressure.
Likewise, because production hydrofractures are under
compression, we again assign a 100 psia pressure at
production well 543N, even though the producers are pumped
off.
For history matching, the pressure and temperature
profiles are found sequentially for each layer. First, the short-
time temperature response, which is caused mainly by heat
conduction, is used to assign a thermal diffusivity to each
layer. Next, the hydraulic diffusivity and the area and shape of
the hydrofracture across each layer are allowed to vary in time
and are iterated until an acceptable match of temperature at all
of the observation wells and the cumulative steam injection is
found, The evolution of formation properties on each side of
IN2U is medeled separately. However, during the first 225
days of injection, the hydraulic diffusivities in corresponding
reservoir layers are constrained to be equal on both sides of the
hydrofracture. For the best match of the temperature response
to steam injection in IN2U, the volumetric heat capacity of
each layer in the steam zone varies between 66.0 and 72.5
BTU/ft3-OF, the thermal diffusivity of the oil zone ranges from
1.32 E-6 to 4.84 E-6 ft2/s, while permeability and hydraulic
diffusivity vary from a low of 0.010 mD and 1,26 E-6 ft2/s to
a high of 0.698 mD and 9.36 E-5 ft2/s.
In what follows, we describe the evolution of the heat
522
injecting part of the hydrofracture in IN2U, the match of the
cumulative injected steam, and the temperature responses in
seven observation wells. Figure 7 illustrates the evolution of
the shape of the hydrofracture in IN2U inferred from the
temperature responses in all of the observation wells and the
cumulative steam injection. Depths on the plot vary from the
top of the diatomite at 800 ft down to 1500 ft, while the
perforated interval of well IN2U is from 1110 to 1460 ft. Zero
on the x-axis indicates the location of IN2U. Positive distances
lie to the north of IN2U and negative distances to the south.
Hydrofracture shapes in the east and west directions reported in
Fig. 7 are, of course, mirror images of one another. The
hydraulic diffusivity of the diatomite rock varies across the
hydrofracture area. This variation is coded in grayscale at the
beginning of each time step shown in Fig. 7. A shading of
black indicates a hydraulic diffusivity of zero, and light gray
shading, a maximum hydraulic diffusivity of 1.0 E-4 ft2/s.
Note that Fig. 7 shows different histories of hydraulic
diffusivities for the eastern and western parts of the pilot.
Figure 8 presents the actual cold water equivalent (CWE)
of steam injected through IN2U, the results of the best history
match, and the volume of steam injected to the east of west of
the IN2U fracture plane from the history match as a function
of the square root of time, Measured pilot injection is given as
a thick dashed line while the predicted injection is a solid line
and agreement is excellent. At least two distinct regions of
injectivity are apparent in Fig. 8. At times less than roughly
17 daysl/2 (289 days) injectivity is quite low, but it increases
dramatically at later times. fIre increase of hydrofracture area at
308 days, shown in Figs. 7e and 7f, is largely responsible for
the increased injectivity. Additionally, Fig. 8 teaches that
significantly more of the injected steam flowed to the west of
IN2U as compared to the east.
The pilot temperature response and history match results
as a function of depth for the eastern portion of the pilot are
given in Figs. 9 to 12. The measured temperature response is
given as a dashed line and simulation results are given as
darkened circles connected by straight lines. Figure 9 displays
formation heating atLO15 and Fig. 10atLO12 between O md
1230 days. Notably, both plots show that substantial heating
of the diatomite occurred between the top of the diatomite at
roughly 800 ft and the top of the perforations at 1100 ft. Thus,
the hydrofracture shapes displayed in Fig. 7 extend some 300 ft
above the perforated interval to 800 ft. Interestingly, the
temperature responses displayed in Figs. 10 to 12 for wells
LO12, MO1, and M02, respectively, show that steam flow
and heating of the southern wing of the IN2U hydrofracture
was delayed for about 500 days of steam injection.
Temperature response was weaker and slower in the southern
wing of IN2U as compared with the northern wing. The
northern wing showed strong heating at times less than 500
days of steam injection with the strongest temperature
responses across the perforated interval.
SPE 35396 A. R. KOVSCE~ R. M. JOHNSTON, AND T. W. PATZEK 5
Figures 13 to 15 give the temperature history match for
the west side of IN2U. Figure 13 which displays the
temperature response for Well LO 14 shows undeniable
evidence of steam flow into diatomite layers above the
perforated interval, consistent with the results in Figs. 9 ad
10. A dramatic temperature response in the D and G cycles
between depths of880and 1120ftgreatly exceeds the response
measured at depths corresponding to the perforated interval.
Take, forexample, the Gcycleresponse between 308 and 560
days. In this time interval, theheating increases roughly by a
factor of 10. At 308 days, the temperature difference is
approximately 25 F while at 560 days it is 330 F. The
temperature data for Well LO13 which is farther from IN2U
and to the west of L014 also shows an exceptionally strong
temperature response above the perforated interval in the D and
G cycles. The maximum temperature increase at LO 13 in the
G cycle between 843 and 1230 days is nearly equal to that in
LO 14. These results indicate very rapid movement of steam
through the G cycle and are evidence for development of a
horizontal feature such as a horizontal fracture or network of
tilted fractures connected to the hydrofracture in IN2U.
Possibly, the hydrofracture which is vertical between 1110 ad
1460 ft turned in the horizontal direction toward 543N in the
process of growing upward. However, this tkacture or network
of fractures does not extend from IN2U to 543N as steam
breakthrough did not occur in 543N until after roughly 3 1/2
years steam injection, Fig. 4. Across the H and the I cycles,
corresponding to the perforated interval, the temperature
response for LO 13 lags behind LO 14; the increase of
temperature between the constant square root of time intervals
is roughly constant indicating relatively little or no evolution
of the formation permeabilityy across those cycles.
Figures 13 and 14 do show some discrepancies between
the actual and model-predicted temperature response between
depths of 900 and 1100 ft. For LO 14 in Fig. 13, the values for
the predicted temperature increase at 843 and 1230 days in the
D and G cycles exceed the actual values by about 40 F. These
differences are caused by either an incorrect estimate of the
pressure profile in these two cycles or by measurement errors
in the temperature surveys. If the pressures across the D and G
cycles are lower than estimated, the corresponding saturatd
steam temperature will be less. On the other hand, the
temperature profiles across the D and G cycles at 843 and 1230
days am nearly vertical suggesting that the particulm
temperature logging tool may be measuring its maximum
temperature. Agreement between the actual and the model-
predicted temperature differences at 1230 days across the D ad
G cycles is much better in LO13.
The temperature history for well LO 11 is displayed in
Fig. 15. In addition to a temperature response above the
perforated interval, we can see strong, relatively early heating
of the formation to the northwest of IN2U, corroborating our
picture of strong, early-time steam flow in the northern wing
of the injection hydrofracture.
Putting aside the massive temperature response to the
west of IN2U above the perforated interval, the IN2U response
was more uniform across the perforated interval than in IN2L.
The temperature response at well LO11 is greater than the
LO 12 response by a factor of 2 to 3, but LO 11 is closer to the
hydrofracture plane and should show more rapid heating. As
shown in Figs. 11 and 15 of Part I, LO11 and L012 display
strong temperature responses in different cycles and much
vertical asymmetry of heating by steam injected in IN2L.
However, here both LO 11 and L012 display vertically
symmetric temperature responses. Across the G, H, and the
top of the I cycles in Figs. 10 and 15, the temperature profiles
have similar shapes. Also, comparing LO15 and LO 14 across
the depths corresponding to the perforated zone, we find
roughly the same degree of heating in both the eastern and
western portions of the pilot area.
Comparison to Microseismic Results
Passive imaging of the seismic energy released during the
hydrofracturing process in IN2U and IN2L [3] indicated that
each hydrofracture &tTered greatly in its initial shape and its
position with respect to the injection well. To compare the
microseismic results with the results presented here, Fig. 16
details the final shape of the IN2U aod IN2L hydrofractures
given in Figs. 7i and 7j with a dark line, while the darkerml
squares give the location of microseismic events obtained by
analyzing the arrival times of shear waves at the observation
wells.
Figure 16a demonstrates that the effective hydrotiactum
area resulting from steam injection in IN2U is greater than the
hydrofracture area indicated by microseismic imaging.
However, the microseismic events are consistent with the
shape of the fN2U hydrofracture found here, especially with the
large extent of the wing to the north of IN2U. Further, there is
little reason to expect that a hydrofracture undergoing
continuous steam injection at pressure gradients approaching
the fracturing gradient would not grow as IN2U apparently has.
Although the results presented here are non-unique, they ate
highly constrained. That is, we have matched both the
cumulative steam injection and the time varying areal
distribution of heat, Both of these quantities depend on
hydrofracture shape and the formation permeability. The
asymmetric wing of IN2U to the north across the G, H, and I
cycles is necessary to simulate a temperature response at MO 1,
but a matching, symmetric wing of equal length to the south
would lead us to over predict injection substantially.
Figures 16a and 7 indicate that the fracture in IN2U grew
upward substantially beyond the region of microseismic
activity. However, temperature surveys in the observation
wells and the cumulative steam injection indicate that the
fracture was contained within the diatomite and did not extend
to the overlying Tuhire formation. A full interval temperature
response plot given in Fig. 4 of ref. [7] does not show any
523
6 ANALYSIS OF HYDROFRACTURE GEOMETRY
SPE 35396
transfer of heat from the Diatomite to the Tulare by either
conduction or convection of steam. The steam mass balance
shown in Fig. 8 further supports this contention, in that our
simulations place all injected steam into the Diatomite while
correctly distributing heat across the pilot area at all time
levels.
Figure 16b teaches that the microseismic analysis of the
high-pressure hydrofracturing of IN2L yielded an initial
symmetric ,penny shape for the lower hydrofracture; however,
the shape after approximately 1000 days of steam injection
was roughly rectangular but still symmetric. In the upper
portions of the hydrofracture between 1500 and 1680 ft, wing
length found from history matching the Phase II steam
injection response and that indicated by the microseismic
events are roughly equal. As compared with the upper portion
of IN2L, the lower wings increased their length many fold.
Between 1850 and 1950 ft, the effective wing length has
increased from 10 to 15 ft up to roughly 100 ft. As in the case
of IN2U, there is little reason to expect that the hydrofracture
would not grow while undergoing continuous steam injection
near the fracturing gradient.
The comparison of microseismic imaging and the
effective heated areas of hydrofractures confirms that
microseismic events are diagnostic of the most active portions
of the hydrofracture and it shows that independent analyses
have yielded similar hydrofracture shapes and orientations for
the Phase H pilot. The hydrofracture in IN2U appeared to be
highly asymmetrical with a wing length around 250 ft,
whereas both analyses indicate that the hydrofracture in IN2L
is symmetrical with wing lengths of roughly 160 ft. Further,
regions of the Diatomite in which multiple microseismic
events were recorded, generally accepted steam more readily
than regions which showed little microseismic activity.
Discussion
Evolution of matrix hydraulic diffusivity also played a
profound role in the temperature response history of the pilot
area adjacent to IN2U. Consistent with the IN2L interpretation
in Part I, hydraulic diffusivities of individual layers on each
side of the hydrofracture were constrained to be equal for the
first 225 days of steam injection. Afhvard, the hydraulic
diffusivity in each simulation layer on either side of the
hydrofracture was allowed to evolve independently.
To the west, as shown by the grayscale shading in Fig.
7, hydraulic diffusivities were relatively low and uniform.
Between O and 308 days, hydraulic diffusivities across the
simulation grid remained constant or changed only slightly, At
308 days, the hydrofiacture extended significantly to the north
as indicated in Fig. 7e. Simultaneously, the shading of the
layers immediately above the perforated interval of IN2U
evolves from a dark to a lighter gray indicating a sharp increase
in the hydraulic diffusivity of these layers. Likewise, the
contrast of layer shading shows that the hydraulic diffusivity of
the region immediately above the perforated interval greatly
exceeds that across the remainder of the hydrofracture. At later
times shown in Figs. 7g and 7i, the shading lightens even
further. We conclude that a horizontal feature such as a fracture
or network of fractures is responsible for the dramatic
temperature increase above the perforated interval and to the
west of IN2U. Note that the extent of the horizontal fracturing,
as indicated by the shading in Figs. 7e, 7g, and 7i, appears to
be confined to the region around the wellbore,
Interestingly, except for the region above the perforated
interval with the dramatic temperature response, the hydraulic
diffusivities for the western portion of the pilot area appear to
be decreasing gradually over time. The shading in Figs. 7e, 7g,
and 7i across the depths 1200 to about 1450 ft darkens because
hydraulic diffusivity decreases slightly.
To the east, matrix hydraulic diffusivity for all
simulation layers remains constant or decrwses gradually as
indicated by the shading of Figs. 7f, 7h, and 7j. There are no
dramatic, or even slight, increases in hydraulic diffusivity. For
example, the shading for layers between 1200 and 1400 ft in
these figures darkens gradually. This indicates that the
formation may be gradually plugging toward the east.
Since increases in hydraulic diffusivity and temperature
response to the west far outweigh those to the east of IN2U,
greater steam injection to the west is expected. Figure 8
indicates that the calculated cumulative volume of steam that
flowed to the west is greater than that which flowed to the east
by a factor of about 3.5. Further, the east and west curves
begin to diverge at roughly 308 days when the west hydraulic
diffusivity first begins to increase.
To contrast steam flow in the formation adjacent IN2U
with that adjacent IN2L, Fig. 17 replots the cumulative steam
injection through IN2L [7], history match cumulative
injection, and the volumes of steam that flowed to the east and
west of IN2L from the history match. Here, steam flow to the
east and west is much more balanced. Initially, there is a
greater volume of steam injected to the west because the in
situ pressure gradient is directed toward producer 543N and aids
westward injection, At roughly 30 daysn, eastward cumulative
injection overtakes westward because there was rapid
propagation of steam along the boundary between the L and
the M cycles toward 543P leading to injector/producer linkage.
The hydraulic diffusivity histories for IN2L in Fig. 9 of
ref. [7], and for IN2U in Fig. 7 allow us to generalize about
the evolution of the diatomite formation during steam
injection in the Phase II pilot. In the absence of horizontal
fracturing which occurred above the perforated interval to the
west of IN2U, or the rapid propagation of steam along the
boundary between the L and the M cycles to the east of IN2L,
hydraulic diffusivity tends to decrease gradually over time.
Since hydraulic diffusivity is proportional to formation
permeabilityy, our model indicates that the Diatomite formation
524
SPE 35396 A. R. KOVSCEK, R. M. JOHNSTON, AND T. W. PATZEK 7
may plug gradually upon years of steam injection unless (i)
dramatic extensions of hydrofracture planes occur, (ii)
fracturing of the rock matrix perpendicular to the hydrofractures
occurs, and/or (iii) the formation is heated substantially
preventing silica precipitation.
The gradual decrease in hydraulic diffusivity over most of
the formation to the west of the injection hydrofracture planes
may be the cause of the decrease of productivity of 543N prior
to steam breakthrough. Recall, Fig. 4 shows that the second
derivative of production with respect to the square root of time
is negative. On the other hand, the productivity of the close
producer, 543P, in Fig. 3 increases even though some
plugging of the formation occurs to the east of IN2U amf
IN2L. Apparently, the large hydraulic diffusivities of the
diatomite matrix opposite IN2L, and the concomitant steam
influx to the east of the central portion of the hydrofracture as
indicated by the lightest shading in Figs. 9f, 9h, and 9j of ref.
[7] are sufficient to give 543P an overall increasing
productivity.
Further examination of Fig. 7 suggests that a sizable
fraction of the steam injected in IN2U flowed above the
perforated interval. To quantify this fraction, Fig. 18 plots the
ratio of the cumulative steam injected above the perforations to
the total cumulative injected steam in IN2U as a function of
the square root of time for our simulations. Between O and 8
days l/2 (64 days), the ratio gradually ramps up to 0.21 ad
then remains level until about 17.5 days 12 (308 days). Over
this time period, the hydraulic diffusivity of all simulation
layers remains approximately constant as indicated by the
grayscale shading in Fig. 4, At 308 days, the hydraulic
diffusivity of the layers immediately around the well and above
the top of the perforations increases sharply; hence, more
steam is injected into these layers relative to layers opposite
the well perforations. As Fig. 17 illustrates, the ratio of steam
injected above the perforated interval to the total injection
climbs to slightly less than 0.50 between 17.5 and 35.1
days 1/2 due to the contrast in hydraulic diffusivity for zones
above and below the top of the perforations. We calculate that
roughly 43% of the total injected steam contributed to heating
of diatomite cycles lying above the well perforations.
Conclusions
The Phase II steamdrive pilot in the South Belridge Diatomite
demonstrated that significant heating of the formation could be
achieved by injecting steam through two separate
hydrofiactures that span the entire 1000 ft reservoir column.
Heating in the pilot area was caused by both convection of
steam and hot condensate and heat conduction through zones of
the reservoir which steam did not sweep. Heating due to
convection was substantial and could not be neglected.
Steam injection totaled 390,000 BS CWE resulting in a
cumulative oil production of 160,000 BO and an estimated
incremental response of 100,000 BO. Further, the total oil
recovery in the pilot was gauged at 9~o of OOIP and the
incremental oil recovery at 6% of OOIP after 5 years of steam
injection.
Increases of steam injectivity for both injectors IN2U ml
IN2L was linked to the growth of the injection hydrofractures
and to increased formation hydraulic diffusivity. Solutions for
the shapes of the injection hydrofractures are highly
constrained by the areal distribution of heat available from
temperature surveys at the 7 observation wells employed in the
pilot and by the cumulative injected steam.
Comparison between hydrofracture shapes obtained in
this analysis and those obtained from locating the
microseismic events that cccurred during hydrofractunng was
favorable. Steam flow occurred in regions with large densities
of microseismic events. However, this comparison
demonstrated that hydrofracture shape during steam injection
evolved dynamically. Hydrofracture wings can, in some cases,
increase their length many fold during steam injection.
Specifically, analysis of diatomite heating caused by
steam injection through the upper hydrofracture which is
perforated over the interval 1100 to 1460 ft revealed:
1.
2.
3.
4.
Heating of the upper half of the Diatomite was highiy
asymmetrical in the horizontal direction. Substantially
more steam flowed into the northern wing of the
hydrofracture, as judged by temperature response, than
into the southern wing.
Horizontally, asymmetric heating was linked to the shape
of the injection hydrofracture for steam injection in both
IN2U and IN2L,
Above the perforated interval and to the west of IN2U,
formation temperature response was extreme and rapid.
This behavior was most likely caused by a horizontal
fracture or network of fractures originating near the
hydrotlacture in IN2U and pointed toward 543N. Since
this horizontal feature has large hydraulic diffusivityy,
much of the injected steam flowed above the perforated
interval of IN2U.
Except for regions of the Phase H pilot that exhibited
rapid heating reminiscent of a horizontal fracture and
prior to steam breakthrough at a production well,
hydraulic diffusivity either slowly decreased or remained
constant in time indicating a slow plugging of the
formation.
Acknowledgments
We thank Shell Western E&P for releasing the injection,
pressure, and temperature log data from the Phase H pilot, ml
also Drs. P. D. Patel and H. J. Vinegar of CalResources,
LLC. and Shell Development Co., respectively, for helpful
advice, We thank PV Technologies for donating computer time
525
8 ANALYSIS OF HYDROFRACTURE GEOMHRY SPE 35396
AND MATRIWFRACTURE INFRACTIONS DURING STFAM INJFCTION
andsoftware for reading and interpreting the raw temperature
data. Additionally, T. W. Patzek acknowledges unrestricted
research grants ffom Unocal.
This work was supported by the Assistant Secretary for
Fossil Energy, Office of Gas and Petroleum Technology, under
contract No. DE-AC03-76FSOO098 to the Lawrence Berkeley
Laboratory of the University of California. This project was
initiated as part of the California Oil and Gas Alliance between
producers and national laboratories and is funded through the
Natural Gas and Oil Technology n Partnership.
Nomenclature
Ac=
K=
M=
P
=
Qi =
t
=
T=
x
=
Xf =
vertical fracture area, ft2
thermal conductivity, BTU/s-ft-F
volumetric specific heat, BTU/ft3-OF
pressure, psia
heat injection rate, BTU/ft3-s
time, s
temperature, F
horizontal distance, ft
steam front location, ft
Greek Letters
a = hydraulic diffusivity, ft2/s
K
=
thermal diffusivity, ft2/s
LT = total mobility, ft2/psi-s
v = total compressibility, psi-1
References
1.
2.
3.
4.
Johnston, R.M. and Shahin, G. T., Interpretation of
Steam Drive Pilots in the Belridge Diatomites, SPE
29621, presented at SPE 65th Western Regional Meeting,
Bakersfield, CA, March 1995.
Kovscek, A. R., Johnston, R. M., and Patzek, A. R.,
Evaluation of Rock/Fracture Interactions During Steam
Injection Vertical Hydrofractures, SPE 35396, presented
at SPE Western Regional Meeting, Bakersfield, CA,
March 1995.
Ilderton, D. C., Patzek, T. W., Rector, J. W., and Vinegar,
H.J., Passive Imaging of Hydrofractures in the South
Belndge Diatomites, SPE 28383, presented at SPE 69th
Annual Technical Conference, New Orleans, LA,
September 1994.
Vinegar, H. J., Wills, P. B., DeMartini, D, C.,
Shlyapobersky, J., Deeg, W. F. J., Adair, R. G., Wwpel,
J. C., Fix, J.E., and SorrelIs, G. G., Active and Passive
Seismic Imaging of a Hydraulic Fracture in Diatomite,
JPT (Jan. 1992), 44(1), 28-34, 88-90.
5.
6.
7.
Marx, J.W. and Langenheim, R, H., Reservoir Heating
by Hot Fluid Injection, Trans. AIME (1959), 216,312-
315.
Schwartz, D. E., Characterizing the Lithology,
Petrophysical Properties, and Depositional Setting of the
Behidge Diatomite, South Belridge Field, Kern County,
California, in Studies of the Geology of the San Joaquin
Basin, S.A. Graham and H.C. Olson, Editor, The Pacific
Section Society of Economic Paleontologists and
Mineralogists, Los Angeles, CA, (1988) 281-301.
Kovscek, A. R., Johnston, R. M., and Patzek, T. W.,
Interpretation of Hydrofracturv Geometry Using
Temperature Transients 1: Model Formulation ad
Verification, In Situ (1995), submitted, 1995.
Table 1. ID Simulation of Oil Production with
Pressure Interference
~[ 543P 543N Parameter
175 330 E-W producer spacing (ft)
300 290 Length of both wings of fracture (ft)
350 850 Pay thickness, w/o Brown Shale (ft)
1215 1450 Average Pay Depth (ft)
I
200(depl)
I
665 I P, (psia)
I
50 50 Producer BHP (psia)
0.19 0.19 Absolute permeability (mD)
0.5 0.5 Porosity (fraction)
30 30 Oil Gravity (API)
30 30 Rock Compressibility
200 4(H3 P-, for compressibility talc. (psia)
50 56 Oil saturation (%)
11 11 Gas saturaion (90)
I
0.33
I
0.37 I WOR for total mobility calculation
1 1.6 I GOR (mcf/stb) for total mobility
calculation
0.35 0.50 Fractional k,,, at which total
mobility is calculated
526
SPE 35396 A. R. KOVSCEK, R, M. JOHNSTON, AND T. W. PATZEK 9
543N
/
< G
100 ft
ig. 1 - Plan view of the Phase 11 steam drive pilot in
ection 29 of the South Belridge Diatomite.
0
0 10 20 30 40
Squara ROOSof Tima on Pmdwtkm Sims 10EKI(Oaya)
Fig. 3- Cumulative oil and water production in 543P.
Diamonds denote analytical solution.
0
0 10
------..1---- --,
20 30 40
Square Root of lime on Injection Since 10/90 (Days)
Fig. 2- Cumulative steam injection and slope in IN2U.
Note that slope of slope is +600 BCWE/day.
I
Inject Steam
Temperature
I
Oil
0 30 60 90
Sqrt(Days on Production)
Fig. 4- Cumulative oil and water production in 543N.
Diamonds denote analytical solution. Steam
injection started st 60 square root of dsys.
527
10 ANALYSIS OF HYDROFRACTURE GEOMETRY SPE 35396
AND MATRDUFRACTURF INFRACTIONS DURING S~
700
800
900
1000
-
...........-. ..-.. -.. .- . . ................... .... .... .. . .. ......
12
-
b
d
8 -
.=
~
v .
26 :
...)..,,.. . ............. ..............
m
2
k
$4
a
~
Q2
----- Actuat Pressure History
j
o
l . . ..l. .l. J . . ..l. J.. l.. . .
F
: ___
-
~
p -
.
,.--... -... -,---
-z Y+.
-.:
.%

..
..4%
-- ...
. ......
G
1
t
..........._.. -.. _~.-
b---! --
:
.........
~- F----33
1200
1300
1400
1500
!
9
. . . . . . . . . . . ..-... -... -. L. . r.+.. ._...-
.,,,,S,,
---
..........
--... *... ..
:.=+.
..
. . . . . ..Porosity L@
-_POrosity for Simulation .--:-.:
. ..-. ..-
-/
.. . .. ......
I
.. . .. .. .. ..
0.4 0.5 0.6 0.7
Porosity
Fig. 5 - Porosity log for Well L015 and porosities
by layer for history match.
o 5 10 15 20 2s 30 35
Square Root of Time (days 12,
Fig. 6. - Steam inJectlon pressure history for IN2U.
8CKt
~ ICOO
g Ilou
~ 1200
1300
1400
lSOQ
II,,,
8CX3
m
~ 1000
- 1100
i
I200
I300
1400
1500
/+,
803
~ Iftoo
- 1100
1
120U
1300
140Q
b
Ir@alllic
dlmsivily
(n% XId )
o 1.()
1S30
I
I I I
-1OOOKX3ZM:
i- w-m N2U ft)
1~
Distancefroy IN2U (ft)
.lCk a 100 200 3
Dis.lancef[gjnIN2U (ft]
-160 0 Iti 260 3
f3iSt,WEC froy IN2U(ft)
ml
m
~ lcoo
; 1100
~ 1200
1300
1400
1500
I!!l!lk
800
~ lm
; IIcm
g 12LIJ
131XI
1403
1500 1
Soo
~ 1000
- 1100
1
120Q
13CS3
1400
I&
1
15C0
I
I I I
.lm o 100 2m 3(
IXstsme fro)m1N2U(IV
r 1
.l&l 6 Kb Z& 3(
Distsnce f[m)m1N2U(ft)
-lb o 160 Z& 3(
DISISI)CC frorl IN2U (ft)
-160 0 160 2ti 31
Distancefm~ IN2U (fl)
800
900
~ lCOO
- llm
!
1200
1300
1400
k
1W3
[
[ 1 f I
-h ii Ilk 2&l 360
-Kk o 100 203 300
Distancefr~jnIN2U(ft)
Distancefy IN2U [ft)
Fig. 7- Side vtew of IN2U fracture zone and hydrtmtic diffusivtdca for history matching (a) wcsti O-25days, (b) em O-25days,
(c) westi 25-308 days, (d) eesti 25-308 days, (e) west30S-547 days, (fl eest:308.547 days, (g) wrst: 547-850 days, (h) esstiS47-850 days,
(i) west: 850-1230 days, (j) eask850-1230 days.
528
SPE 35396 A. R. KOVSCEK, R. M. JOHNSTON, AND T. W. PATZEK
11
200 . . . .
Upptk Inj~ctor:IN2~ 1 1
100 .
5
0
0 5 10 15 20 25 30 35
Square Root of Time (days 2,
Fig. 8 - Steam injection history for IN2U including steam
injection to the eastern and western portions of the pilot.
700
800
900
1000
1300
1400
1500
Observation Well L015 ~
. . .. .. .. .. .. .. ..... ..... . ... .. .. .. .. .. .. .. .. .. .. .. .. .. .. ,-----
;... .. .. .. .. .. .. .. .. .. .. .. ... .. .. .. .. .. .
... .. .. .. .. .. .. .. .. .. .. .. j.. ..----- .. .
0 50 100 150 200 250
Temperature Increase (F)
Fig. 9 - Temperature survey (dashed lines) and history
match (symbols and solid lines) results In Well L015.
700
[ I I I
Observation Well LOl~ g
I
800 L.-*W
. .... .................... .... .... ..................... .... ...
J
... ............... .... .... .... .... .......p..._.. _..
d
-Ri~i
*-.s - i
>. c]
...-.
900
D
.
1000 :.+: i .................... .... .... ..................... .... ...
~ .:
>, -. ~*. G
~1 loo . + - :
I
--- y .
a
.:
\
:.
-. :-.;
.:- ;
1200
H
2s9 d 559 d S43 d i ~
1
is.--.=%.
....-... -...+...............i....-... -...4
::wd I 11
...................~ ...._. _..
. . . .. .... ............... ..... .... .... ................ .... .... ...
1500
0 50 100 150 200 250
Temperature Increase (F)
Fig. 10 - Temperature survey (dashed lines) and history
match (symbols and solid lines) results in Well L012.
700
1 I I I
Observation Well MO1 ~
~
J
800
.4-,. --.;+-... . . . . . . . ..y---- . O-. . . y. !-. . . - . . . . . p- . . -, . . .
c:
..............................-... -... +...-... ...............-... -
900
D
1000
.... .... .... .................... ..+...-. . .. ........... .... ....
~
G:
~....-. ..-. ... .. .. .... .. .. . +...-...-...-
%
n
H;
-+-...-... -.-. +-. -... - .....~....-. ..-... -
1
14004-
~
...... .... .... ..............:....-. -.-... +..._...-----j...._..._.-. #
1500
i 1 1 i
1
0 50 100 150 200 250
Temperature Increase (F)
Fig. 11 - Temperature survey (dashed lines) and history
match (symbols and solid lines) results in Well MO1.
529
12 ANALYSIS OF HYDROFRACTURE GEOMETRY
SPE 35396
AND MATRDUFRACTURF INFRACTIONS DURING STFM INJFCTION
700
1 1 I I
Observation well MOj 1 i
800
I
... .. ... .. .. . . .. .. .. .. ... . ... .. ... .. .. .. . .. .. .. .. .. .. .. .. .. .. .
c
1
900 L-""-"-""-"-'-:"-" -"-"-"-'"+-""-"-""-"-"-':"-"-"-"-"-' -:-"-"-"-"--3
-vii ~
D
*.
1000 . ......w._.._...- ....c.. ............-... -..<-...-..~_...-...--
-:
*:
\
.
.-+ . . .
G
---- .
...--...*,.4............-.. !-...-... -..i ---------- 1 I*.. ....
f;
. ... .... .................... .. . .... .... .... .... .... .
i
...........+ . .... ....... ........ .+.. .- ...+...- ,----
0 50 100 150 200 250
Temperature Increase (F)
Fig. 12 - Tern erature survey (dashed lines) and history
f match (symbo s and solid lines) results in Well M02.
7oo~. . . . . . . . . ..c . . . ...1 l!... . . l. . [ . . . . . . . . . J
~Observation Well LO14 ! ~
800 .Vq<:
.. .. .. .. .. .. .. .. . .. .. .. .. .. .. .+_.. m_.,_-
-~ $~ ]
.-
. - --
c:
900 a %:.-$ ------- 2:=-----
.
:-- --~-~ D
1000 ,.-
,..- +..- .... .... ..
4-
: .-=. - ...-. .--h .- ...-.
--+.-t
n
~
= 1100 --mw,:.;:-$:... -
,..-, -.+. .-.. .. .... ..
~ -. ##. -
! 308;d .- >-;;
{
n 1200 . 32 : 1 84~~~*:
H
: .
<.
*
..-..! -560 d.... -~.
-H- -----+-------
1300 - :
/
:----
,Olq, ,f
;230d :
1,:
I
!# --i
1400 1. f-* ; ::; ;
l;
:-+;
, .,._..L -.-+.. -... ~.. *.. -... ----- ...-*. -_...- .... ..
*: ---*..-
1500
0 100 200 300 400
Temperature Increase (F)
Fig. 13 - Temperature survey (dashed lines) and history
match (symbols and solid lines) results In Well L014.
700
, ,
Observation Well L013 ;
80 O& . .. .. .. .. .. .. . .. .. .. . .. . ..+ .. . .. . .. .. .. .. .. .. .. .. .. .. .. .
c
-. -.--. --.-*.--. --... ....................I . :-..
900 \
---%
----- :
--- +...
.
-. --:
---
~1
D
--:. .
1000 -...::.:+. -.-+-:-2 - -% --g.% .--...
~
..:
~---- :
.=--,---- - G
4-. G
s1100 560 d..843 ;~~;j--j :
~
/{
. . .... ...+.. -... -... -...
-.
1200 ! $-230 d j
,
,: ,1 :
1
..: .... . . -.+. ---------- ..+-...-. ... ...
1300
1
p---w - .
.-.-+ ----- ----+ ---- -_... +..+ .. .. .. .. .. .. .
1500
1. . . . . . . . . 1. ..
0 100 200 300 400
Temperature Increase (F)
Fig, 14 - Temperature survey (dashed lines) and history
match (symbols and solid lines) results in Well L013.
700
1 # 1
Observation Well LO1l ! i
800
900
1000
~
~ 1100
~
Q 1200
1300
1400
1500
0 100 200 300 400
Temperature Increase (F)
Fig. 15 - Temperature survey (dashed lines) and history
match (symbols and solid lines) results in Well LO1l.
SPE 35396 A. R. KOVSCEK, R. M. JOHNSTON, AND T. W. PATZEK 13
1000 & . .. .. .. .... ....+ ..... ..... .. .. .. .. .. .. .. ... .. ....~
1300
1400
>-... -,..- .... ..................... .... .... ..............
-1oo 0 100 200 300
Distance From IN2U (ft)
Fig. 16a - Comparison of IN2U fracture shape from
history matching at 1230 days and microseismic events.
1400
..... .... .... ... .... .... ...--.:. -... - .... .... .... ...--------
..... .... ...- ,.+...-..-. ..--.; .-... - ...-... -.*.-. ... .... . ..
2000
-200 -100 0 100 200
Distance from IN2L (ft)
Fig. 16h - Comparison of IN2L fracture shape from
history matching and 1050 days and microseismlc events.
1
Lower Injector: IN2L ~
Y ~
E
150
[
.........
~
C%.-
u&
~-
~ :100
..........
:Z
4
Sb
*
s-
6
----= Actual Injection
50
z
O 5 10 15 20 25 30 35
Square Root of Time (daysi)
Fig. 17 - Steam injection history for IN2L including steam
injection to the eastern and western portions of the pilot.
0.6 , 1 r
- Upper Injector: IN2U ~ ~ ~
:; ::; :
,..,..... ...........;.,..,,.,,.,:.,..,..,,.,:,,..,..,,..:..................,,.,.,
0.4 -;:::
:........................y........-
0.3 -
;;; ;
::: :::
::: ::
0.1 -
;;; ;;;
;;; ;;;
;:: ::: :
0
0 5 10 15 20 25 30 35 40
Square Root of Time (days 2,
Fig. 18 . Fraction of cumulative steam injected above
the perforated interval of IN2U.
531