Вы находитесь на странице: 1из 17

SPE-184008-MS

Assessment of the Areal and Vertical Sweep Efficiency in Cyclic Carbonate


Reservoirs of the Middle East - A Case Study from a Mature Field

Ruaa Abdalla, ADCO; Jorge Gomes, ADNOC; Mohammed Al Kobaisi, PI; Gehad Mahmoud, ADCO

Copyright 2017, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Middle East Oil & Gas Show and Conference held in Manama, Kingdom of Bahrain, 6-9 March 2017.

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 have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Carbonate reservoirs contain almost half of the world's hydrocarbon proven reserves. Because of the
horizontal and vertical heterogeneities present in cyclic carbonate reservoirs, more than 50% of the oil in
these rocks is trapped or bypassed and yet to be recovered, even after water injection. This fact by itself
motivated researchers to investigate ways to improve the recovery factors by implementing new recovery
schemes coupled with advanced well designs and completions. This paper aims to analyze the value of
adding ICDs (In Flow Control Devices) in horizontal wells to improve the areal and vertical sweep efficiency
and the recovery factor.
A high resolution mechanistic model was built for a large carbonate reservoir in X-Field, with the
aim to capture the small scale sedimentary and diagenetic cycles, including the dense stylolitic intervals.
Geostatistical algorithms were implemented to propagate the petrophysical properties within a sequence
stratigraphic framework. The 3D static model was built based on real field data, using several wells.
However, for the purpose of the simulation work done in this study, only a crestal oil producer and a
downflank water injector in a line drive system with different completion strategies is analyzed. Different
sensitivities were then studied in order to investigate the impact of various reservoir descriptions and
multiple well completions, involving vertical wells, standard horizontal wells and horizontal wells with
ICDs, on the overall sweep efficiency and recovery factor.
This work showed that changing the fluid injection/production rates as well as completion intervals is
not the only controls to achieve higher recovery factors. By implementing smart completions with ICDs
and proper well placement, higher volumetric sweep efficiencies were achieved, resulting in an incremental
gain of 9% recovery factor.

Introduction
The carbonate reservoir, where the model was built, belongs to the Kharaib Formation which
chronologically belongs to Lower Cretaceous Epoch. It was deposited in the Barremian-Aptian ages. The
reservoir's source rock is Diyab Formation and it's sealed by an impermeable unit between Shuaiba and
Kharaib Formations.
2 SPE-184008-MS

The reservoir is divided into two main zones, a lower zone with an average thickness of 60 ft and an upper
zone with an average thickness of 100 ft. It extends over large areas of the Middle East and deposited in a
carbonate ramp system which was primarily governed by 2 main transgressive-regressive cycles, Figure1.

Figure 1Generalized lithostratigraphic section across the Arabian Peninsula and Iran (Alsharhan and Nairn, 1986)

This carbonate reservoir is thus a vivid example of vertical reservoir heterogeneity where the upper zone
has high permeability and the lower zone has considerably lower permeability. The reservoir cyclicity and
various petrophysical properties lead to uneven hydrocarbon sweep between the injector and the producer.
Fluid flow prefers low resistance paths and therefore the water front has the tendency of pushing the oil
faster from the high permeability zones bypassing large volumes of hydrocarbon in the lower zones.
The work presented here aims to refine a method of enhancing the sweep efficiency and improving
the recovery factor by the use of smart well completions with a careful consideration of well spacing
and completion design. The 3D static model was built using Petrel software and the flow simulation was
conducted with Eclipse 100. The model was built with the grid dimensions of 2525 meters in the X-Y
plane and with a total of 72 layers in the Z-direction with 1 ft vertical resolution. The total number of cells
after upscaling was 1.2 million. This mechanistic static model helped in capturing the small 4th order cycles
and the vertical heterogeneity within this carbonate reservoir.
The flow simulation was done using different development strategies; initially, simulation was run on
a pair of a vertical injector and producer where injection/production rates and perforation intervals were
varied. Horizontal wells were then added to the model to enable accessibility to the unswept hydrocarbons
in the lower zone. The scenarios tested were ranked based on the volumetric sweep efficiency and recovery
factor. Finally, ICDs were added in the optimum well configuration to further improve the sweep and
increase the recovery factor.
The main goal of this paper is to optimize the sweep efficiency in order to increase the recovery factor
by leaving less bypassed oil behind. The three main factors that affect recovery are the EA (Areal sweep),
Ev (Vertical Sweep) and local displacement efficiency. The areal sweep considers the ratio of area that
has been swept to the total pay zone area, while the vertical sweep deals with the swept height and its
relativeness to the total height of the reservoir. The areal sweep is highly affected by the mobility ratio
which is defined as the ratio of the mobility of injectant (displacing fluid) to the mobility of reservoir
hydrocarbon (displaced fluid). Besides, areal sweep is also affected by the injection rate and pattern, and
rock and fluid properties. Having a mobility ratio of less than 1 is considered favorable (Guliyev, 2008). A
SPE-184008-MS 3

favorable mobility ratio will result in a stable and even sweep to the reservoir fluid leaving less bypassed
oil. In general, when the displacing fluid is much less viscous than the displaced fluid, viscous fingering can
occur and the displacement becomes unstable resulting in a poor uneven sweep and low recovery factor.
This research aims to deploy smart completion with various well segmentations to have a better control on
the sweep efficiency. Several flow simulations were performed in order to achieve a judicious integration
of displacement efficiency together with proper well spacing and completion design. This was evaluated
with a 3D mechanistic model by integrating the micro and macro scales honoring the geology.
Several completion strategies were tested in order to increase the sweep efficiency and recovery factor
and reduce the water cut. ICDs work on reducing the heel-toe effect which is caused by the uneven pressure
drop along the horizontal section due to reservoir heterogeneity. In injector wells, ICDs are added to have
better control on the injection rates which will help in optimizing the sweep and increase the profit gained
from the well (Fernandes et al., 2009). In high permeability zones, the ICDs apply high pressure allowing
small amounts of fluid to enter the formation, while in low permeability zones the ICD backpressure is
minimized in order to let more fluid enter the formation.

Description and Application of Equipment and Processes


To conduct the flow simulation, a high resolution mechanistic static model was built with data sets provided
by the operating company which included 3D seismic, well logs, well markers, digitized core description,
conventional core analysis and MICP (Mercury Injection Capillary Pressure Data). Furthermore, PVT and
SCAL data were subsequently added to the simulation model.
Upon importing the wells into Petrel, a quality control test was conducted on the data provided from 16
wells in order to check for logs availability. Wells with core and log data were used for correlation and to
build the structural and stratigraphic framework of the static model. Furthermore, fine zonation and layering
were done; the main well logs used were GR, density, and porosity. 18 surfaces were created using 19
horizons. Subzonation/layering was done based on vertical variation on the petrophysical properties and the
thickness of each zone. This process enabled the capturing of the dense stylolitic intervals, and was then
compared with the digitized core analysis.
In order to propagate the properties in the inter-well region, a stochastic approach was used, Sequential
Gaussian Simulation (SGS) for porosity modeling and Sequential Gaussian Simulation with co-kriging for
permeability. SGS is a stochastic approach that reproduces the spatial heterogeneity of the propriety by
utilizing specified variograms. This modelling approach revealed to be more realistic than the standard
deterministic methods based on moving weighted average techniques; it was able to capture the small
heterogeneity of the reservoir while honoring the well data.
Multiple realizations were created to capture the extent of uncertainty for each model. The main controls
of the stochastic simulation technique are the spatial correlations determined by the variogram models used.
In this study, the parameters used for the isotropic variograms were provided by the operating company
and no special nested variograms or anisotropy was implemented. Table 1 shows the parameters used for
porosity and permeability, respectively.

Table 1Adjustment of parameters for SGS Simulations

Parameter Model/Values

Variograms (porosity)
Exponential Spherical
Variogram (permeability)

Nugget 0

Major Range (m) 4000

Minor Range (m) 4000

Vertical Range (ft) 10


4 SPE-184008-MS

To establish an understanding of the flow zones, rock typing was generated. The process of rock
typing was done based on porosity vs permeability cross plots, capillary pressure curves, and the SMLP
(Stratigraphic Modified Lorenz Plot). Capillary pressure curves were generated from MICP data. MICP data
for 19 core plugs were provided; grouping of the rock types was based on the drainage capillary pressure
characteristics together with the pore throat size distributions. The Pc curves plotted were then grouped
based on the entry pressure, shape of Pc curves and the Swirr (Irreducible Water Saturation) (Frank et al.,
2005; Gomes et al., 2008; Aliakbardoust and Rahimpour-Bonab, 2013; and Rushing et al. 2008). Four rock
types were established as presented in Figure 2.

Figure 2Capillary pressure curves for four different rock types

To start with saturation modeling, J-functions for each rock type was generated after obtaining saturation
values for the different rock types using the following relations:
(1)

(2)
where a, b and are fitting parameters, and Sw is the corresponding saturation value. Different fitting
parameters were implemented and the best fit was established using the parameters in Table 2.
SPE-184008-MS 5

Table 2Fitting parameters for each rock type

RRT1 RRT2 RRT3 RRT4

a= 1.00 1.30 2.00 1.82

b= 0 0.02 0.02 0.03

= -0.55 -0.6 -0.6 -0.46

The Leverett-J Function along with wettability studies were also used to confirm the rock typing. The
height above free water level was determined by the following relationship

(3)

where is the porosity, is the surface tension, w and o are the water and oil density respectively. A
summary of the values used are shown in Table 3.

Table 3Parameters used in the study

Constants

cos = 32.6 dyn/cm w = 64.13 lbm/ft3

K = 50 md o = 39.95 lbm/ft3

= 0.25

Another simplified technique for rock typing used is the classical porosity vs. permeability cross plots; the
petrophysical properties obtained from core and log data are grouped into clouds that correspond to different
rock types (Figure 3). The Pc curves established earlier where then assigned to the different rock clouds.

Figure 3Porosity vs permeability cross plot

Imbibition relative permeability curves for each rock type are given Figure 4. The parameters needed for
the trend of the relative permeability of each rock type were derived from the SCAL data of the 19 core
plugs provided by the operating company.
6 SPE-184008-MS

Figure 4Relative permeability curves for each rock type

Validation of Reservoir Zonation with Lorenz Plots


After building the static model and before initializing the simulation model, validation of the petrophysical
zonation was performed using the Stratigraphic Modified Lorenz Plot. Flow capacity (k*h) versus
normalized storage capacity (*h) was plotted for the entire reservoir section. The number of slopes in the
Lorenz Plot (Figure 5) indicates the number of flow units in the reservoir. It also quantifies the contribution
of the possible flow from each layer (Izgec, 2012). From the Stratigraphic Modified Lorenz plot, it was
interpreted that zone XIIIU contributes to 60% of the flow and only 15% of the storage capacity; this gave
an idea as to where to place the producers and injectors in the model.
SPE-184008-MS 7

Figure 5Stratigraphic Modified Lorenz Plot

Flow Simulation
The static model was loaded into Eclipse, and subsequently PVT and SCAL data were added together with
the well completions; the fluid properties of the reservoir and the rock physics model were inserted and
various development strategies were explored using a Black oil model.

Vertical wells
A pair of vertical wells was used in the model; an injector and a producer; perforations were added in
different intervals (reservoir zones) for each of the wells. The perforation was set to be 6-7 perforations
per foot. The scenario that gave the best result in terms of areal and vertical sweep is summarized below
(Figure 6):

Perforations in XI to XII for the producer

Perforations in XIV to XV for the injector

Rates: 3000 bpd for the injector and producer for 4 years
8 SPE-184008-MS

Figure 6(a) Perforations at the producer and injector. (b) Areal sweep and (c) Vertical sweep

In order to quantify the sweep efficiency for this scenario, the number of grids that had saturation
changes was counted and compared to the total number of cells. Another scenario where both injection and
production taking place in the best flow zone (XIIIU) was simulated with oil and water rates being 2000
bpd for 4 years. Figure 7 indicates that most of the contribution is from the upper part of the reservoir.

Figure 73D window showing the (k/phi) model on the left and the sweep
with the perforations done in XIIIU for both the producer and the injector

Figure 8 depicts the results of varying the rates and/or completion intervals. It can be seen that the
highest sweep efficiency was achieved when the vertical producer is completed in XIIIU while the injector
completed in the lower part of the reservoir.
SPE-184008-MS 9

Figure 8Sweep behavior for different flow rates and completion intervals

Horizontal wells
Next we explore the impact of horizontal well completions. Different sensitivities on the well trajectories
and completions were introduced to the model. The length of the horizontal wells was kept constant at 3000
feet and the following parameters were varied:

The trajectory of the wells (inclined versus horizontal)

Completion strategy for the producer-injector pair

A summary of some of the scenarios (eight in total) is shown in Table 4. In all of the cases, water was
injected into the lower reservoir section. The scenarios were ranked in an ascending order from #1 to #8
based on the sweep efficiency.
10 SPE-184008-MS

Table 4Ranking of the scenarios; Incl refers to the inclined section of the well and Horiz to the horizontal part

The scenario that showed the best sweep efficiency is shown in table 5, where the horizontalization for
the injector was done in the lowest zones (sub-zones 11-18) while the producer was horizontalized in (sub-
zones 7-8).
SPE-184008-MS 11

Table 5Best scenario settings

Figure 9 depicts the production profile of the best case scenario. It can be seen that the plateau of
production, 3000 bpd, was sustained for 5 years. At the end of the fifth year the water cut increases, oil
production starts to decline and oil rate drops to 1500 bpd and then gradually decreases to a low of 500 bpd
by the end of the simulation run of 30 years.
12 SPE-184008-MS

Figure 9Oil Production rate (left axis) and water cut (right axis) for the best case scenario

Horizontal wells with ICDs


To further enhance the recovery factor, well segmentations with passive ICDs were added to the model. The
number of nozzles per ICD was decided based on the relationship between the rate and the pressure drop
displayed in Figure 10. For the purpose of this study, 6 nozzles per ICD were added.

Figure 10Flow characteristics for an ICD (Aadnoy and Hareland, 2009)

The ICDs were added to both injector and producer. The horizontal well section was divided into
compartments based on the lateral variation of petrophysical properties. For both the producer and injector
in the 3000 ft long horizontal section, 6 ICDs with six nozzles per ICD were simulated (Figure 11).
SPE-184008-MS 13

Figure 11Set-up of ICDs in Petrel

Different completion strategies with the addition of ICDs and various sensitivities were studied such
as flow coefficient, nozzle inner and outer diameters and space between the nozzles. It was noted that
by considerably increasing the ICDs flow coefficient and cross sectional area and decreasing the nozzle
spacing, the sweep became more uniform and the recovery factor started to increase. Figure 12-C and
12-D show the water saturation profile with four compartments established with one to two ICDs per
compartment.
14 SPE-184008-MS

Figure 12Water saturation profiles at the end of the simulation run; (A) water saturation profile for
the base case displaying the grids with saturation changes; (B) water saturation profile with ICDs
added to the producer; (C) and (D) depict water saturation profiles at the end of the simulation run
by gradually increasing the inner diameter for compartment 3 from 44 mm to 66 mm, respectively

Furthermore, the addition of ICDs led to a longer plateau period and higher recovery as compared to the
best case scenario without ICDs as shown in Figure 13.

Figure 13Oil production rate (left axis) and water cut (right axis) compared to the base case (dashed line)
SPE-184008-MS 15

Figure 14 shows the water cut behavior for the different scenarios tested.

Figure 14Water cut comparison

In order to quantify the zonal contribution across the horizontal section, simulated PLT (Production
Logging Tool) was generated, with and without ICDs; the change in the slope shows the incremental change
in the contribution (see Figure 15). It can be inferred that adding ICDs to the model gave accessibility to
the lower zones that were not contributing earlier.

Figure 15Simulated PLT (A); PLT profile with ICDs (B); PLT profile without ICDs (C)

There is no discrete way of quantifying the sweep efficiency. Visually the sweep efficiency could be
monitored on the basis of the waterflood frontal movement. For the purpose of this study, the grids with
16 SPE-184008-MS

no saturation changes were first filtered out with the use of property calculator (Figure 16). Afterwards,
the recovery factor for each simulation run was calculated using volumetrics. The overall sweep efficiency
is a function of the vertical and areal sweep efficiencies and the microscopic displacement efficiency. The
provided microscopic displacement efficiency for the carbonate reservoir in study has a range between
70-85%.

Figure 163D domain showing the filtered results for saturation changes

For the assessment of the areal sweep, a layer by layer view was displayed and the number of grids per
layer that had saturation changes was compared to the total number of grids per layer. The arithmetic mean
was then calculated using the below equation:

(4)

Likewise, for the vertical sweep efficiency the saturation was then tracked section by section in the K-
direction and the average of the saturation change was calculated by the following:

(5)

The recovery factor calculated from the resultant areal and vertical sweep was (43 to 49%) based on the
(70-85%) range of microscopic displacement efficiency provided. A 9 % increase in the overall recovery
factor was attributed to adding ICDs to the horizontal well completions.

Conclusions
This paper investigated the use of ICDs as a mean to enhance the volumetric sweep efficiency and recovery
factor for a given carbonate reservoir exhibiting substantial vertical heterogeneity due to the depositional
and diagenetic cyclicity. This was achieved by modeling a pair of wells (injector and producer) in a direct
line-drive pattern with different completions, including smart completion with ICDs and well placement
strategies. As a result of adding ICDs in an optimized configuration, the recovery factor was noticeably
improved by about 9%.
SPE-184008-MS 17

Acknowledgement
Portions of this work have been taken from the MS thesis of Ruaa Abdalla to fulfill the research requirement
for the MS degree in petroleum engineering at the Petroleum Institute. The authors would like to thank Abu
Dhabi Company for Onshore Petroleum Operations (ADCO) for providing the field data and permitting us
to publish the results.

References
1. Aadnoy, B.S. and Hareland, G., 2009. Analysis of inflow control devices, Offshore Europe. SPE-
122824-MS
2. Alsharhan, A. and Nairn, A., 1986. A review of the Cretaceous formations in the Arabian
Peninsula and Gulf: Part I. Lower Cretaceous (Thamama Group) stratigraphy and
paleogeography. Journal of Petroleum Geology, 9(4): 365391.
3. Aliakbardoust, E. and Rahimpour-Bonab, H., 2013. Integration of rock typing methods for
carbonate reservoir characterization. Journal of Geophysics and Engineering, 10(5): 055004.
4. Drakeley, B.K., Douglas, N.I., Haugen, K.E. and Willmann, E., 2001. Application of reliability
analysis techniques to intelligent wells, Offshore Technology Conference. Offshore Technology
Conference. OTC-13028-MS.
5. Fernandes, P., Li, Z. and Zhu, D., 2009. Understanding the roles of inflow-control devices in
optimizing horizontal-well performance, SPE Annual Technical Conference and Exhibition.
Society of Petroleum Engineers. SPE-124677-MS
6. Frank, S., Narayanan, R., Hansen, P., Allen, D.F., Albrechtsen, T., Steinhardt, H., Raven, M.J.,
Fordham, E., Bize, E. and Rose, D.A., 2005. Carbonate rock typing using NMR data: a case
study from Al Shaheen field, offshore Qatar, International Petroleum Technology Conference.
International Petroleum Technology Conference. IPTC-10889-MS
7. Guliyev, R., 2008. Simulation study of areal sweep efficiency versus a function of mobility ratio
and aspect ratio for staggered line-drive waterflood pattern, Master of Science Thesis, Texas
A&M University, Texas.
8. Gomes, J.S. and Alves, F.B., 2013. The Universe of the Oil and Gas Industry: From Exploration
to Refining. Partex Oil and Gas.
9. Gomes, J.S., Ribeiro, M.T., Strohmenger, C.J., Naghban, S. and Kalam, M.Z., 2008. Carbonate
reservoir rock typing-the link between geology and SCAL, Abu Dhabi International Petroleum
Exhibition and Conference. Society of Petroleum Engineers. SPE-118284-MS
10. Izgec, O., 2012. Understanding waterflood performance with modern analytical techniques.
Journal of Petroleum Science and Engineering, 81: 10011
11. Rushing, J.A., Newsham, K.E. and Blasingame, T.A., 2008. Rock typing: Keys to understanding
productivity in tight gas sands, SPE Unconventional Reservoirs Conference. Society of Petroleum
Engineers. SPE-114164-MS

Вам также может понравиться