You are on page 1of 11

SPE 53133

Waterflooding of A Heavy Oil Marginal Reservoir


H. H. Hanafy, SPE, Gulf of Suez Petroleum Company, and A. M. Mansy, SPE, Gulf of Suez Petroleum Company

Copyright 1999, Society of Petroleum Engineers, Inc.


This paper was prepared for presentation at the 1999 Middle East Oil Show held in Bahrain,
20-23 February 1999.
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
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
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
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300
words. Illustrations may not be copied. The abstract should contain conspicuous
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.

Abstract
This paper shows how Gupco (Gulf of Suez Petroleum
Company) has successfully developed a heavy oil marginal
reservoir by using high technologies such as reservoir
simulation, 3-D visualization, coiled-tube dual completion, and
coiled tube pipeline. The July 53 block was discovered in
January 1986 with the drilling of the vertical platform proving
well J-53. The well was drilled to test the Lower Rudeis
formation in a 90-acre sliver block to the west of the main
Lower Rudeis July field. The test indicated a potential of 1350
BOPD with an API gravity of 19 degrees, an oil viscosity of 7
CP, and a gas-oil-ratio of 85 SCF/STB. For this reservoir fluid
type, the recovery factor was expected to be less than 5%
under primary conditions. Therefore, the development of this
fault block was put on hold until further studies could
determine the optimum plan of depletion. Starting in 1993,
GUPCO began studying small reservoirs such as the J53 block
with a PC-based black oil simulator. In the case of J53, the
simulator was used to match the short term DST from the
discovery well, and then create multiple depletion scenarios.
Waterflooding, it was discovered, would provide the best
development plan. Use of the simulator provided insight into
the recovery process, and reduced the development risk.
Under primary recovery, the simulator confirmed a low
ultimate recovery of 3% of the OOIP. With the optimal
waterflooding plan, however, recovery could be improved to
23% of the OOIP. Considering that normal Lower Rudeis oil
recovery approach 54% of the OOIP under waterflooding (in
the main field with 32 degree API oil), the simulator provided
a much more conservative estimate. The development of this

fault block was started in May 1996 by drilling a high angle


producer designed to be parallel and close to a major fault to
encounter the full Lower Rudeis section. Waterflooding was
started in December 1996 by completing an existing Nezzazat
producer with a unique coiled-tube dual water injection - oil
production completion. A 3 1/2 inch coiled tubing line was
laid to provide injection water from a near platform at almost
7000 BWPD. Performance to date indicates a successful
waterflood project. The wells production rate has increased
from 1300 BOPD, before waterflood start, up to the peak
production rate of 4100 BOPD.

Background
The July 53 block was discovered in January 1986 with the
drilling of the vertical platform proving well J-53. The well
was drilled to test the Lower Rudeis formation in a sliver block
to the west of the main Lower Rudeis producing field. The
Lower Rudeis formation was drillstem tested in well J53. The
test indicated a potential of 1350 BOPD with an API gravity of
19 degree, an oil viscosity of 7 CP, and a gas-oil-ratio of 85
SCF/STB. The lower gravity and higher viscosity crude as
compared to the main Lower Rudeis crude and the RFT
pressure data which showed a virgin reservoir pressure both
indicate the J53 block is a separate fault block.
Calculations of recovery factor expected under primary
conditions from this fluid type is less than 5% and
development was put on hold until further study could be made
to determine the most effective way to drain this 90 acres
block. Therefore, the well J53 was temporarily abandon for
future reentry.
Subsequently, two more wells penetrated the fault block,
J92 and J93 to further define the block. The well J92 is a mud
line suspended well located in 122 ft of water depth about 150
ft from the J15 platform. The well did not tied-back to the
platform J15. The well J93 was drilled from platform 37 and
was completed as a Nezzazat producer.
None of the three wells have encounter the full Lower
Rudeis section as oil bearing either due to faulting or due to
penetrating the oil-water contact. Therefore, none of the
existing wells can deplete the whole Lower Rudeis section.

H. H. HANAFY, A. M. MANSY

The Lower Rudies in block J53 is a marginal reservoir with


only 21 million barrels of oil originally in place. For all of the
above mentioned reasons, the development of such marginal
reservoir would be too risky.

Simulation Objective
A multidiscipline team study was successfully helped to
identify an optimum plan of depletion for this sliver fault
block. The objective of this study was directed toward using a
PC-based black oil simulator to estimate the potential for
waterflooding the Lower Rudeis. Several water injection
schemes have been tested with this simulator in order to
minimize the potential risk of development.

Reservoir Data
Structure. Efficiently tapping additional reserves in the
mature, structurally complex Gulf of Suez oilfields has
improved significantly since applying todays digital, 3-D
modeling and visualization tools.
Reevaluation and
exploitation of the areas remaining reserves has been
conducted by multidisciplinary teams using state-of-the-art
workstations and geologic modeling software chosen
specifically for its ability to unravel structural complexity from
scattered subsurface data.
The Gulf of Suez exploitation team actually tracked the
time required to image Suez fault blocks using new 3-D
technology vs. traditional 2-D cross section and mapping
methods. They found that the time it took to produce a
rigorous model of a fault-bounded block was decreased by a
factor of four or five.
The geospatial modeling technique incorporated in Earth
Vision, 3-D modeling package selected by Gupco, provided
the tools needed for verifying and modifying interpretations.
Fault surfaces were modeled, intersected, and displayed in 3D, allowing July asset team to understand the complex fault
shapes and intersections.1
Fig. 1 shows that the July 53 block is controlled by two
faults, the B1 to the West and D North to the East and South.
The D North fault is well controlled while the B1 West fault is
less controlled. The oil-water contact, determined by RFT and
log data, is at -8650 ft subsea which is about 900 ft shallower
than the oil-water contact of the main Lower Rudeis reservoir.
The structure is mapped at the top of the Lower Rudeis
sand which have three penetrations, wells J53 , J92, and J93.
The structure dips to the northeast direction with the strike of
the structure trending southeast to northwest. This will have a
significant effect on water displacement during water injection
as will be shown later.
The structure cross-section shown in Fig. 2 shows how the
Lower Rudeis formation is laying between the two parallel
faults B1 and D. None of the existing wells penetrate the full
Lower Rudeis section as oil bearing. Therefore, to optimally
deplete the Lower Rudeis reservoir, the path of a new well

SPE 53133

should be designed to intersect the top of the structure near the


B1 fault and drill parallel to the B1 fault in the updip location.

Production Data. No production has been established from


the Lower Rudeis in the July 53 block. The short term DST
performed in the discovery well was the only available
production data to be matched by the simulator. The DST was
performed at two different production chocks and both
indicated an average production rate of 1350 BOPD with an
API gravity of 19 degrees and a GOR of 85 SCF/STB.

Pressure Data. The RFT data for the three wells J53, J92 and
J93 all indicated that the July 53 block is isolated from the
main Lower Rudeis reservoir. The RFT data plot presented in
Figure 3 shows the following:
1. An oil gradient reflects a low API gravity of 18 degrees
compared with 32 API gravity for the main Lower Rudeis
crude.
2. A water-oil contact of -8650 ft subsea which is about 900 ft
shallower than the main Lower Rudeis contact.
3. An initial reservoir pressure of 4330 psi at reservoir datum
of - 8600 ft subsea. This is too close to the main Lower
Rudeis initial pressure (4380 psi) when they are correlated
to the same datum.
4. Although the six years apart between the discovery well J53
and the subsequent wells J92 and J93, no pressure
depletion was indicated.
5. No water injection response was indicated due to injecting
water in the main Lower Rudeis

Fluid Properties. Two bottom hole oil samples were collected


during drillstem testing of the Lower Rudeis formation in well
J53. The PVT analysis was conducted at two different labs
and the results were in generally good agreement except the oil
viscosity.
To select which sample can be used for further reservoir
evaluation, an AMOCO computer program was used to quality
check the lab analysis. This program uses the AMOCO`s
modified Redlick-Kwong equation of state (ARKES) to
predict the fluid properties at both reservoir and separator
conditions from its composition. The compositional analysis
for one of the two samples was in error, therefore, only one
sample was subjected to quality check.
Due to large content of heavy ends (64 mole %), some
equation of state parameters and component properties for the
heavy ends pseudocomponents have been subjected to change
to fine tune the predicted properties with the lab
measurements. A good match between the lab and predicted
properties was obtained by reducing binary interaction
parameters between methane and heavy ends and by reducing
both critical temperature and critical pressure while increasing
critical volume for heavy ends pseudocomponents.

SPE 53133

WATERFLOODING OF A HEAVY OIL MARGINAL RESERVOIR

Table 1 compares the ARKES predicted properties before


and after tuning with the lab measurements at the bubble point
pressure. Fig. 4 shows, as an exaample, an excellent match
between the lab and predicted oil formation volume factor at,
above, and below the bubble point pressure.
The J53 area oil has a fairly low API gravity of 18 degrees
with a BPP of 765 psi. The bubblepoint Bo, Rs, and viscosity
values are 1.182 BBL/STB, 161 SCF/STB, and 7.43 CP,
respectively. The low API gravity for the J53 sliver block was
also indicated from the well head samples collected during the
DST of well J53 (19 API). The RFT oil gradients of the three
wells drilled in this block (J53, J92, and J93) also reflected a
low API gravity of 18 degrees (Fig. 3).
This oil is significantly heavier than the main Lower
Rudeis field oil which have a higher API gravity of 32, a
higher BPP of 2222 psi, a higher Bo of 1.411 BBL/STB, a
higher Rs of 588 SCF/STB, and a lower viscosity of 0.65 CP.
The large difference in the oil properties between the J53
sliver block and the main field Lower Rudies is one of the
evidences that reflect the isolation of this block.
Adjustment of the PVT Data for Field Separator
Conditions. The reservoir process is simulated in the
laboratory by a differential liberation process (DVA). The
separation of oil and gas at lower temperature and pressure
encountered in the surface separators is not accounted for in
the DVA experiment.
At surface separator conditions, some intermediate
components which are in the gas phase at reservoir
temperature and pressure may condense at surface conditions
and appear in the tank liquid. The effect of this larger amount
of tank oil is to reduce both Bo and Rs. The effect of
separator conditions is very important as the oil volatility
increases; the corrections can approach 20% (even more for
volatile oils). The adjusted values are the appropriate ones for
use in material balance calculation. If differential liberation is
used instead of the flash liberation at surface separator in the
OOIP calculations, this can lead to errors of 10 to 20% or
more.2
The July field is tested offshore at the separation
facilities located at the July 10 platform. The field test
separator is operating at 70O F and 50 psi. The separator
recovery option of the PVT simulator was used to calculate the
bubble point Bo and Rs at the above mentioned field separator
conditions. The program shows that the Bo and Rs at the BPP
are 1.091 BBL/STB and 95 SCF/STB, respectively. This
represents a reduction of 8% and 44%, respectively in the
corresponding DVA values.
Correction to Bo and Rs at pressure other than the bubble
point pressure are best made by adjusting the DVA values by
an amount equal to the correction previously determined for
the bubble point oil. This result in Bo and Rs curves which
are parallel to the DVA data, but offset down by the correction
evaluated at the bubble point pressure. A method has been
proposed for accomplishing this adjustment mathematically.3
This method is dependent on differential and flash laboratory

measurements of Bo and Rs at bubble point pressure.


Equations used in this adjustment are as follows :
ofb
o = od
odb
ofb
and, Rs = Rsfb - (Rsdb - Rsd)
odb
The flash laboratory measurements are represented here by the
results of separator recovery option of the PVT simulator. The
adjusted DVA curves as a function of reservoir depletion are
presented by the lower curves in Fig. 5.
The Impact of Oil Properties on the Depletion Strategy of
J53 Sliver Block. The low oil compressibility (7.5 X 10-6 psi1
) led to a low primary recovery estimate of only 5% of the
OOIP assuming that the completed well can contribute to flow
at a low reservoir pressure equal to the BPP of 765 psi. Due to
the low primary recovery estimate, the development of this
sliver block could not be justified and the well J53 was
temporarily abandoned for future re-entry.
The high viscosity for J53 oil of 7.43 CP is significantly
much higher than the main July Lower Rudeis oil viscosity of
0.65 CP. The relatively high viscosity is uncommon in the
Gulf of Suez (GOS) where oil viscosity in most reservoirs is
less than 1.0 CP. GOS reservoirs in which secondary recovery
projects have been initiated have all contained low viscosity
oil. Although, the J53 oil has a higher viscosity than most
GOS reservoirs, it is not much higher than secondary recovery
projects in other parts of the world. A review of the major
waterfloods in West Texas shows that most reservoirs have
viscosities in the range from 2-5 CP. By analogy, a secondary
recovery factor of about 20% could be expected for the J53
sliver block. Secondary recovery factors for normal GOS
crudes are within the average of 45%.

Rock Properties. While a lack of core data hindered


development of a reservoir description for the July 53 block,
other analogous Lower Rudeis core cut from the July J6-6 well
in the main reservoir was available for use in this study.
Combined with existing log data, a satisfactory reservoir
description was developed.
The rock permeability was derived from the log-core
correlation. Both rock porosity and water saturation data was
collected from the open hole logs of the three wells J53, J92
and J93. The July J6-6 Lower Rudeis core provided the
relative permeability data for July 53 block in the simulation
study. This data reflect the criteria typical of water-wet rock
in which the residual oil saturation to water would approach
21%.

Layering. The July main Lower Rudeis sand was subdivided


into seven geological layers namely, from top to bottom, RLS7
through RLS1. This sand is under-lain by a thick shale layer

H. H. HANAFY, A. M. MANSY

namely RLS20. For the purpose of reservoir modeling, the


RLS5 in block 53 was subdivided into two layers upper and
Lower. Figure 6 shows how these layers are distributed along
the three wells J53, J92, and J93. As shown from this Figure
the well J93 is oil bearing in the top three layers, well J53 is
oil bearing in the middle two layers, while well J92 is oil
bearing in the bottom three layers. Therefore, none of the
three wells is oil bearing in the whole Lower Rudeis either due
to faulting or penetrating the oil-water contact. Also, J93 is
the only well in block 53 that encountered the majority of
Lower Rudeis except the bottom layer RLS1. As will be
indicated later by the model, layer RLS1 contributes to the
lowest OOIP when compared to the other layers. Therefore,
well J93 can play a very important role in the development
plan of block J53.
Parameters for different layers are in the ranges listed
below:
Horizontal permeability (Kh)
Porosity
Water saturation
Average Gross thickness
Average Net thickness
Net/Gross thickness

16.0
0.13
0.20
29.0
14.0
0.28

64.0 md.
0.18
0.44
174 ft.
141 ft.
0.91

Vertical permeability was assumed to be only one percent of


the horizontal due to shale streaks between different layers.
The relative permeability data, normalized to the initial water
saturation of each layer, was provided to the simulator.

Model Development
Simulation of block 53 was accomplished with the PC-based
MORE reservoir simulator. MORE is a black oil simulator,
and it was loaded on an IBM 486/66 MHz computer with 16
Mbytes of RAM. Plots of fluid saturation changes were
created with Tecplot, a 3-D graphical program that interfaces
with MORE.

3-D Model. The 3-D Model covered the entire July block 53
structure area as shown on Figure 1. Areal dimensions for the
model are 4100 ft wide by 7628 ft long, for a total simulated
area of 718 acres. Grid block size is 195 ft by 200 ft, resulting
in 22 I-direction blocks and 39 J-direction blocks, for a total
6864 blocks in 8 layer system. Based on the mapping, this
model contained an OOIP of 21.2 MMBO.

History Match
The simulator was used to match the short term DST from the
discovery well. To achieve the DST match, changes to the
most uncertain areas of the reservoir description were made
first, leaving the more accurate measured data such as log
parameters, porosity and pay thickness, unchanged. A trial

SPE 53133

and error process eventually achieved an excellent DST match


by increasing the log-derived permeability four times.

Predictions
The black oil simulator was used to predict the reservoir
performance and oil recovery under both primary depletion
and waterflooding. Waterflood predictions have been made
under different scenarios to reduce the potential risk of
development.
Before doing any predictions the potential rate for each
well in block 53 was calculated by the simulator at the normal
BHFP of July Lower Rudeis wells (2800 psi) which are
producing at open choke with gas lift. The initial match was
performed at the DST BHFP of 3200 psi due to testing the
well at small production choke of only 3/4 or 1 inch without
gas lift.

Primary Depletion Cases. Predictions under primary


depletion were performed at the following model limits:

Shut wells in when oil production declines to 200


BOPD.

Shut wells in when FBHP declines to 1000 psi.


Two cases of primary depletion were tested as follows:
Tie-Back J53 and J92 and Recomplete J93. As shown in
Fig. 6, the three wells together can act as a one well. The
simulator predicts a primary recovery of only 0.74 MMBO
from the three wells which represents 3.5% of the OOIP. Due
to the low primary recovery estimate and the high cost
required for placing the existing wells on production, the
block development was not economically justifiable. Well J53
needs to be equipped with subsea completion. Also, it needs
coiled tube pipelines for oil and gas to transmit and lift its
production to the nearest platform. Well J92 needs a guyed
caisson system to place the well on production with short
length (150 ft) coiled tubing flow lines and gas lift line
between the well and J-15 platform. Well J93 needs to be
recompleted from the producing Nezzazat to the Lower
Rudeis.
Recomplete J93. The primary recovery in this case was
calculated to be 0.56 MMBO which represents 2.7% of the
OOIP. It was decided that the recompletion of J93 to Lower
Rudeis production to be deferred until the remaining Nezzazat
reserves (0.3 MMBO) is produced.

SPE 53133

WATERFLOODING OF A HEAVY OIL MARGINAL RESERVOIR

Secondary Recovery Cases. The simulation study indicated


that the production under primary depletion should not be
permitted because it will lead to a sharp drop in reservoir
pressure. Waterflooding after depleting the reservoir pressure
would then require the injection of large water volume that is
not available. Also, large injection water volume would
require a large size water injection line that will reduce
significantly the economics of the water injection project.
Three waterflood schemes were tested by the simulator
using the following common assumptions:
Minimum BHFP of 2800 psi,
Shut well in when oil production declines to 200 BOPD.
Shut well in when WOR reaches 10.
Case 1: Convert J93 as Injector and Drill an Updip
Producer. As mentioned before, well J93 intersects the
majority of Lower Rudeis section except the bottom layer
RLS1. The simulator indicates that the RLS1 represents less
than 1% of the total Lower Rudeis OOIP (Table 2). So, well
J93 can be used as a down-dip injector to inject water into the
majority of Lower Rudeis section.
To receive the feedback of water injection, an updip
producer should be drilled. The well is intended to penetrate
all productive zones (RLS7-RLS1) to maximize recovery in
the simulator, the well path was designed to intersect the top of
the structure near the B1 fault and drill parallel to the B1 fault
in the updip location (Fig. 2). The well will have a high angle
of 76 degrees. The imaged area shown on the structure contour
map (Fig. 7) shows the bottom hole location of the
recommended well on the bottom layer to be injected RLS2.
This plan yields a field oil recovery of 4.84 MMBO (23% of
the OOIP) at a water injection rate limit of 8000 BWPD.
Fig. 8 represents the predicted oil, water and gas
production profiles with two subsequent water shut-offs to
keep the well on production below the WOR limit of 10. As a
sensitivity run, a 16000 BWPD yields a similar resource
recovery but at a higher rate. The field life in this case will be
9 years instead of 11 years for the 8000 BWPD case.
However, supplying a 16000 BWPD would require a higher
cost water injection line.
Analysis of 3-D oil saturation plots of the reservoir
indicated that the injector J93 acts as an updip injector where
oil saturation changes down dip J93 toward the aquifer. Oil
saturation then changes toward the new producer (withdrawing
point) leaving unswept area between J93 and the new
producer. The structure dips to the northeast direction with the
strike of the structure trending southeast to northwest yield a
gravity force that result-in this performance. Table 2 shows
the breakdown of layers movable oil saturation left by the life
end.
Case 2: Drill an Injector and Drill an Updip Producer.
The simulator was then used to study the effect of a new
downdip injector to inject water into the aquifer and produce
from an updip producer. The location of the new well was
selected to intersect the full Lower Rudeis section between the

two parallel faults. The well angle would be 65 degrees. The


well targets on top RLS7 and RLS2 are shown in Fig. 7.
In this case the simulator predicts almost the same oil
recovery as in Case 1. However, this case would be more
expensive than Case 1 due to drilling of a new injector. Also,
the model indicates that four water shut-offs would be required
to continue producing the well below the given model
production limits (Fig. 9).
Two reasons can explain why the downdip injector does
not improve the oil recovery over the Case No. 1. First, a part
of the injected water may be lost into the aquifer rather than
going up to displace oil. Second, the structure shape creates a
gravity force that prevents water from displacing oil in the
adjacent area to J93 and displacing oil mainly toward the
withdrawing point.
This was indicated from the 3-D
visualization of oil saturation.
Case 3 : Drill an Injector, Drill an Updip Producer, and
Recomplete Well J93. As a trial to improve the areal
sweep efficiency, the model assumes the recompletion of J93
to be a Lower Rudeis producer in addition to the updip
producer. The 3-D plots of oil saturation showed a gradual
decrease in oil saturation change toward the well J93 due to
gravity forces. The model predicts a 1.1 MMBO to be
recovered by J93 and 4.5 MMBO to be recovered by the new
producer. The total field ultimate recovery of 5.6 MMBO
represents 26.4% of the OOIP. Although this case indicates
0.8 MMBO incremental recovery over the two other cases, this
case still less economic than the first case. This case would
require the drilling of a new injector and additional injection
capacity to provide water to support two producers. In
addition, one water shut- off in well J93 and three subsequent
water shut-offs in the new well would be required to enable the
two wells to produce below the model production limits (Fig.
10). As a result, the recommended case is to convert J37-93 to
injection and drill an updip producer from the platform J-15.

Facility Design And Economics


In order to start the waterflood project in the Lower Rudeis of
block 53 without waiting for the depletion of remaining
Nezzazat reserves (0.3 MMBO), a dual completion of well 3793 was chosen. The dual completion involves water injection
through a 2 3/8 inch short string down the backside into the
Lower Rudeis. Production from the Nezzazat formation is by
gas lift gas injection via a concentric string of 2 3/8 inch
production tubing and 4 1/2 inch tubing to contain the gas.
The dual completion well design is shown in Fig. 11. A
conventional dual completion with two tubing strings does not
provide adequate injection rates. The cost of the dual
completion was $ 0.7 MM.
A 3 1/2 inch coiled tubing line was laid from the J-10
platform to the July 37 platform to provide injection water at a
rate of 7000 BWPD. This line was dynamically laid from a
workboat equipped with hydraulic reeling equipment. Refer to
reference 4 for additional details on GUPCO's workboat based

H. H. HANAFY, A. M. MANSY

coiled tubing experience. The cost of the coiled tube water


injection pipeline was $ 0.85 MM. The cost of the
recommended J15-97 producer was estimated to be $ 4.1
MM. However, the final cost of the producer was reached $
5.2 MM to penetrate the formation after faulting out at a
higher angle than the prognosis of 85 degrees. So, the total
project cost was $ 6.75 MM to recover 4.84 MMBO.

Post Appraisal
The production performance of well J15-97 is presented in
Fig. 12. The J15-97 well was completed as a Lower Rudeis
producer in May, 1996 and initially produced at a rate of 9460
BFPD and 4% water cut. This accurately agrees the predicted
model rate of 10,000 BOPD. Due to the lack of the coiledtube production string, waterflooding in the J37-93 well was
delayed till December of 1996. During this period, the J97
production was continually dropped until it reached 1300
BOPD before waterflood get started. Under waterflooding,
the J97 production has increased up to the peak of 4860 BFPD
with 16% water cut in July, 1997. A total of 2.0 MMBO have
been produced from the Lower Rudeis versus 3.5 MMBW
injected till October of 1998. As indicated from the model
runs, water shut-offs would be necessary to enable the well to
recover the estimated reserve.

Conclusions
Gupco as a pioneer company in Egypt, has used state-of-theart technologies including: 3D visualization, 3D reservoir
simulation, coiled-tube dual completion, coiled-tube pipeline,
and multidisciplinary team approach to put the marginal J53
block on production after 10 years of discovery.
Waterflooding is successfully improve the recovery of the 18
API gravity oil in this fault block. The simulation study
indicates that the waterflood ultimate recovery will be about
23% of the OOIP versus only 3.5% by primary depletion.

SPE 53133

Nomenclature
BHFP
= Wellbore bottom hole pressure, psi
= Separator flash liberation oil formation volume
od
factor at reservoir pressure other than the
bubble point pressure, BBL/STB
=
Differential
liberation oil formation volume
odb
factor at the bubble point pressure, BBL/STB
= Separator flash liberation oil formation volume
of
factor at reservoir pressure other than the
bubble point pressure, BBL/STB
=
Bubble
point pressure, psi
BPP
=
Oil
density,
Gm/CC
o
=
Drillstem
test
DST
= Differential liberation analysis
DVA
= Gas-oil-ratio, SCF/STB
GOR
= Original oil-in-place, STB
OOIP
OWOC = Original water/oil/contact, ft
RFT
= Repeat formation tester
R.F.
= Recovery Factor, %
Rsd
= Differential liberation gas-oil-ratio at reservoir
pressure less than the bubble point pressure,
SCF/STB
= Differential liberation gas-oil-ratio at the bubble
Rsdb
point pressure, SCF/STB
= Separator flash liberation gas-oil-ratio at
Rsf
reservoir pressure less than the bubble point
pressure, SCF/STB
= Separator flash liberation gas-oil-ratio at the
Rsfb
bubble point pressure, SCF/STB
= Residual oil saturation
Sor
= Oil saturation at the life end
Sof
= Connate water saturation
Swc
= Water saturation at the life end
Swf
= Ultimate Recovery, MMBO
U.R.
= Oil viscosity, CP
o

Acknowledgments
The authors wish to thank both GUPCO and AMOCO Egypt
Oil Company for their permission to publish this paper. The
authors also wish to express their appreciation to all Gupco
members shared in both the study stage and the project
implementation stage.

References
1. Day, R. A., and Hoffman, K. S.: "Gupco More Efficiently Taps
Suez Oil Using 3-D Modeling," World Oil (September 1997)
47-53.
2. Moses, P.L.: "Engineering Application of Phase Behaviour of
Crude Oil and Condensate Systems," JPT (July 1986) 716, 722,
723.
3. Reudelhuber, F.O., and Hinds, R.F.: "Compositional Material
Balance Method for Prediction of Recovery from Volatile Oil
Depletion Drive Reservoirs," JPT (January 1957) 19-26, Trans.,
AIME, 210.

SPE 53133

WATERFLOODING OF A HEAVY OIL MARGINAL RESERVOIR

4. Hoffman, J.G. et al.: "Reeled Pipelay Cost Reduction Using


Workboat - Based Installation," SPE 29786, MEOS (March
1995).
5. Hanafy, H.H.: Simulation of the July J53 Fault Block With a
PC-Based Simulator, SPE Egyptian Section, the 2nd Petroleum
Computer Conference, Cairo, Egypt, April 1995.

TABLE 1 PREDICTED OIL PROPERTIES VS. LAB MEASUREMENTS @ BUBBLEPOINT PRESSURE


Property @ BPP

Before Tuning

After Tuning

Lab

BPP
o
Rs
o
o
API

952
1.187
193
26.87
0.972
-1.89

765
1.182
169
7.41
0.847
18

765
1.182
161
7.43
0.841
17.8

TABLE 2 REMAINING OIL SATURATION AFTER WATERFLOOD CASE NO. 1


Layer #

Swc, Fraction

Swf , Fraction

Sof , Fraction

OOIP, MMSTB

U.R., MMSTB

R.F., %

RLS7
RLS6
RLS5U
RLS5L
RLS4
RLS3
RLS2
RLS1

0.27
0.21
0.32
0.41
0.44
0.20
0.40
0.39

0.43
0.46
0.53
0.55
0.59
0.55
0.63
0.62

0.36
0.33
0.26
0.24
0.20
0.24
0.16
0.17

2.376
6.244
3.278
4.101
3.050
1.510
0.429
0.157

0.348
1.570
0.772
0.840
0.579
0.476
0.129
0.040

14.6
25.1
23.6
20.5
19.0
31.5
30.0
25.5

21.14

4.77

22.6

Total
Sor
Sof

= Residual oil saturation = 0.21


= Oil saturation at the life end = 1 - Swf - Sor

jjkijjhyguuuuuuuuuuuuuuuuuuyguigyiyugiygiygiyugiyyiygihguihgjhgjhgijfgytftytxfcgfvvvvvvvvvvvvvhghghgh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhggggggggggggggggggggggggggggicccccccccc
cccccccccccccccccccxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxzzzzzzzzzzzzzio

Fig. 1 - J53 Block L.Rudeis Structure Map

H. H. HANAFY, A. M. MANSY

Fig. 2 - Proposed Well Path

Fig. 3 - RFT Data in J53 Block

SPE 53133

SPE 53133

WATERFLOODING OF A HEAVY OIL MARGINAL RESERVOIR

Fig. 4 - Oil Formation Volume Factor Match

Fig. 5 - Oil Formation Volume Factor Adjustment

F/O = Faultedout

Fig. 6 - Lower Rudies Layers Distribution

10

H. H. HANAFY, A. M. MANSY

OW
OC

Fig. 7 - J53 Block L.Rudeis Structure Map

Fig. 8 - Waterflood Prediction - Case 1

Fig. 9 - Waterflood Prediction - Case 2

SPE 53133

SPE 53133
11

WATERFLOODING OF A HEAVY OIL MARGINAL RESERVOIR

Fig. 10 - Waterflood Prediction - Case 3

Fig. 11 - J93 L.Rudeis & Nezzazat Dual Completion

Fig. 12 - J 97A Production Performance

You might also like