Professional Documents
Culture Documents
Saudi Aramco
Journal of Technology
On the Cover
Mohammed A. Al-Ghazal, a Saudi Aramco engineer, runs a highprofile acidizing operation using high-pressure/high temperature
(HP/HT) equipment to enhance the deliverability of a deep, deviated
gas producer.
P R O D U C T I O N C O O R D I N AT I O N
Sami A. Al-Khursani
Ashraf A. Ghazzawi
DESIGN
Samer S. AlAshgar
Manager, EXPEC ARC
CONTRIBUTIONS
Relevant articles are welcome. Submission
guidelines are printed on the last page.
Please address all manuscript and editorial
correspondence to:
EDITOR
William E. Bradshaw
The Saudi Aramco Journal of Technology
C-86, Wing D, Building 9156
Dhahran 31311, Saudi Arabia
Tel: +966-013-876-0498
E-mail: william.bradshaw.1@aramco.com.sa
Abdulaziz M. Judaimi
Khalid A. Al-Falih
Zuhair A. Al-Hussain
Ibraheem Assaadan
Mohammed Y. Al-Qahtani
Charles T. Kresge
Essam Z. Tawfiq
Ali H. Al-Ghamdi
Chief Petroleum Engineer
Abdullah M. Al-Ghandi
General Manager, Northern Area Gas Operations
Salahaddin H. Dardeer
Manager, Jiddah Refinery
ISSN 1319-2388.
COPYRIGHT 2014
A R A M C O S E R V I C E S C O M PA N Y
ALL RIGHTS RESERVED
No articles, including art and illustrations, in
the Saudi Aramco Journal of Technology,
except those from copyrighted sources, may
be reproduced or printed without the
written permission of Saudi Aramco. Please
submit requests for permission to reproduce
items to the editor.
The Saudi Aramco Journal of Technology
gratefully acknowledges the assistance,
contribution and cooperation of numerous
operating organizations throughout the
company.
Spring 2014
Saudi Aramco
Contents
10
29
37
44
53
60
Journal of Technology
ABSTRACT
Acid fracturing or matrix acidizing is often required for increased
hydrocarbon production and long-term well deliverability from
the massive Khuff carbonate gas reservoir in Saudi Arabia, the
holder of the worlds fourth largest gas reserves. Open hole
multistage technologies have demonstrated superior performance in maximizing reservoir contact and productivity through
better distribution of acid across the formation matrix, full interval matrix contribution and efficient propagation of fracture
networks to bypass formation damage and optimize near wellbore conductivity.
The Khuff structure is a late Permian age heterogeneous carbonate sequence that underlies the massive Ghawar field in the
Eastern Province of Saudi Arabia. The Khuff reservoir is subdivided into four separate intervals (A through D), with production
coming mainly from the B and C intervals. Since its initial appraisal in the late 1970s, the majority of Khuff development has
been focused in the relatively more prolific Khuff-C formation,
where coiled tubing acid wash and single-stage acid treatments
were repeatedly performed and evaluated. Over the past five
years, multistage acid fracturing has been implemented in Saudi
Arabias Khuff-C development. The results were carefully evaluated for each trial, and this is now the predominant Khuff-C
stimulation technique.
Up until the middle of 2011, the vast majority of Saudi Arabias horizontal Khuff carbonate gas wells were drilled along
the direction of maximum horizontal in situ stress (max). This
was primarily to enhance wellbore stability and achieve the best
possible penetration rates. Early multistage fracturing treatments in the Khuff generated mostly longitudinal fractures
propagating parallel to the wellbore or in the max direction.
Since then, a holistic approach toward the application of open
hole multistage technology for tight reservoir development has
been adopted.
The complex workflow of this approach calls for, among
other requirements, changing the lateral section placement
strategy and planning the horizontal section to be drilled along
the minimum horizontal in situ stress (min) direction as opposed
to the previous mode of planning along the max direction. Accordingly, understanding the reservoir stress profile, orienting
the horizontal wellbore with respect to the dominant horizontal
2
SPRING 2014
stress component and calibrating the stress profiles against actual open hole logs became the most important highlights of the
new workflow. Radical improvement of stage integrity, multiple
fracture signatures and enhanced well productivity were among
the most important results achieved in developing the deep,
tight gas-bearing zones of the Khuff carbonate reservoir.
Still, an innovative approach was required to address the
more prolific zones of the Khuff-C formation where efficient
matrix acidizing was sought as an alternative to acid fracturing
in wells that could only be drilled in the max direction. Therefore, a purpose built open hole multistage technology system
one that was developed around the idea of distributed limited
entry for placement of matrix acidizing treatments was identified and carefully evaluated.
This article presents the details of the successful application
of this new limited entry, multiple injection technology for optimized matrix acidizing of carbonate horizontal wells, including
trial testing qualification, candidate selection, system design,
functionality, operation and ultimate production profiling.
INTRODUCTION
Tight gas, low permeability reservoirs present a tremendous
challenge with respect to effectively completing and draining
a target reservoir. Cased hole and open hole completions in
horizontal wells offer a cost-effective means of accessing the
entire lateral section, assuming the target pay can be effectively
stimulated. While most open hole completions possess more
advantages than cased hole completions, the challenge with
open hole completions, compared to more conventional cased,
cemented and perforated completions, is understanding and
controlling the fracturing fluid flow into the near wellbore area
of the reservoir.
The Khuff reservoir development has offered opportunities
for a wide array of completion techniques to be implemented and
evaluated, ranging from single stage vertical wells through multilateral wells to multistage horizontal wells. As the focus gradually shifted to tighter parts of the reservoir, the well completions
underwent a process of increased complexity, from verticals to
multilaterals and finally to horizontal multistage fracturing
completions, which are also gaining popularity at the world
scale as the industry taps more unconventional resources.
Early applications of open hole multistage completion technology in Saudi Arabia started in 2007 when the specified number
of stages were run with a typical configuration of a single fracturing port between each two mechanical open hole isolation
packers1-5. Acid fracturing was conducted in the multiple stages
through selective activation of the fracturing ports. It was observed
that in wells drilled in the maximum horizontal in situ stress
(max) direction, the first stage fracture will grow longitudinally,
along the wellbore, parallel to max, causing the potential risk
of overlap with subsequent induced fractures due to natural
fractures and formation fissures. Initiation of the second and
third fractures therefore became a challenge, due to possible
pressure communication across the first induced fracture. To
avoid this fracture overlap, it was decided that wells needed to
be drilled in the minimum horizontal in situ stress (min) direction, allowing transverse fracture initiation perpendicular to the
wellbore. The results from wells with open hole multistage
completions showed increased initial production and less departure from the theoretical hyperbolic decline curves6.
For wells designated for matrix acidizing stimulation, an issue
rose regarding the ability of the single fracture port per stage to
achieve uniform and effective stimulation of the entire stage
length. To maximize the stimulated reservoir volume and reduce
the likelihood of localized treatment of more prolific sections, it
was necessary to think of a better way to address the specific
needs for ensuring efficient stimulation and guaranteeing homogeneous distribution of the acid treatment across the stage
length. This article discusses the first successful application of
an innovative matrix acidizing technology in the more prolific
zones of the Kuff carbonate gas reservoir.
SPRING 2014
is best suited for matrix acid treatments in prolific and naturally fractured carbonate formations. Unlike standard open hole
multistage completion systems9-12, where there is only one
fracture port per stage, the limited entry system features multiple jet nozzles placed in a single interval to create a strong matrix acidizing effect throughout the entire open hole interval13,
Fig. 1.
Stages are created using multiple shear-activated stimulation
jets that are spaced out in the sections of interest and isolated
by hydraulically set mechanical open hole packers, Fig. 2. The
jet nozzles are adjusted and placed according to the reservoir
characteristics determined from open hole logs, enabling controlled injection and leak off for effective flow of the acid treatment into the entire section of the interval. This effectively
places the designed treatment at an optimal rate and pressure
along the stage length, maximizing the development of complex
conductive flow channels, also known as wormholes, throughout the entire stimulated reservoir length.
Each stage consists of a drillable cutter assembly pinned into
a shear housing assembly. Downhole of the shear housing are
shear-activated stimulation jet assemblies, spaced out with casing/liner at predetermined depths. Above the lowermost packer
in each stage is the locking/landing sub. The locking/landing
sub provides isolation of this stage from lower stages and locks
the drillable cutter to prevent it from rotating during milling
operations. For effective setup of the system in the reservoir
section, the liner and the annulus are isolated.
The liner isolation is achieved by the activation balls as they
land on their respective seats in the cutter assembly, closing off
the stages below. Multiple stages can be run, starting with the
smallest activation ball and ending with the biggest ball size at the
top. This mechanical diversion, combined with an advanced
chemical diversion system, allows uniform, precise fluid placement.
Isolation of the annulus is achieved using open hole packers.
The criteria for selecting a packer is to identify which packer
Fig. 4. Acid stimulation through a limited entry, multiple injection matrix acidizing
system for effective wormhole creation.
Fig. 2. Shear-activated stimulation jets and packers in the limited entry, multiple
injection matrix acidizing system; open hole log data are used to position the
packers and refine the fracture stage interval lengths.
SPRING 2014
Fig. 5a. First stage treatment from a single port at the bottom.
Fig. 5b. Second and third stage treatments through multiple stimulation jets.
CANDIDATE WELL
The candidate well, Well-A, was a flank well drilled parallel to
the max direction. After analysis of the open hole log, a threestage limited entry, multiple injection matrix acidizing system
was specified with a hydraulic fracture port for the first stage
and six limited entry stimulation jets for each of the second
and third stages, Figs. 5a and 5b. The system was successfully
deployed to total depth in the 4,000 ft thick gross pay, and
matrix acidizing was pumped as per schedule for the three
stages. As shown in Fig. 6, the opening and closing of each interval and the pumping of the acid treatment went as per design without any operational issues. Table 1 presents some of
the main treatment parameters15.
Figure 7 presents the wellbore layouts as well as the stress
and porosity profiles of Well-A and three offset wells in the
Khuff reservoir. The three offset wells Well-B, Well-C and
Well-D are dual-lateral, deviated and vertical wells, respectively. The candidate well, Well-A, has a 1,300 ft net reservoir
contact, laterally drilled in the max direction. Table 2 presents
some of the reservoir and well characteristics as well as stimulation parameters for all wells. Each of the wells has been drilled
and stimulated with different techniques; however, the reservoir
flow capacity and permeability thickness product (kh) of the
wells are comparable.
Fig. 6. Pumping pressure and rate chart for the Well-A acid treatment.
Stage-1
Stage-2
Stage-3
8,022
8,910
8,210
35
39
41
23
30
30
Fig. 7. Bottom-hole location and trajectory of Well-A and the three offset wells in
the Khuff reservoir.
Open hole multistage technology has been implemented in various fields across Ghawar field to enhance productivity from
moderate to tight reservoirs and to assess the technical and
economic feasibility of this enabling technology in each field
and each reservoir. Figure 8a highlights the distribution of
open hole multistage technology applications in various deep
Well Name
Well Type
Treatment Type
Reservoir Net
Height (Reservoir
Contact) ft
kh
md-ft
Number of
Stages
Average
Pump Rate
(bbl/min)
Acid
Volume
(kgal)
Well-A
Horizontal
Open Hole
Limited Entry,
Multiple Injection Matrix
Acidizing
70 (1,300)
55
25
125
Well-B
2 Laterals
Horizontal
Open Hole
Coiled Tubing
Matrix
70 (1,550)
60
126
Well-C
60 Cased Hole
Perforation
Acid Fracturing
+ Diverter
70
55
34
45
Well-D
Vertical Cased
Hole
Perforation
Acid Fracturing
65
60
34
65
Table 2. Pumping and reservoir parameters for candidate Well-A and its offset wells
SPRING 2014
Fig. 10a. Fold increase in PI between conventional and open hole multistage
completed wells.
Fig. 8a. Distribution of open hole multistage stimulation technology applications
in the deep gas development program in Saudi Arabia.
Fig. 10b. Fold increase in PI of limited entry, multiple injection matrix acidizing
over standard matrix acidizing.
SPRING 2014
pressure, and put on production. Figure 9 presents the normalized productivity index (PI) for the candidate Well-A and the
three offset wells, showing the higher gas contribution from
Well-A.
With the appropriate selection of candidate wells, treatments
with both open hole multistage fracturing and limited entry, multiple injection matrix acidizing completions showed successful
results, and each method contributed to high well productivity.
Figure 10a presents the fold increase in well PI from the application of multistage fracturing over single-stage vertical fracture
treatments. Figure 10b presents acidizing treatments where the
application of limited entry, multiple injection matrix acidizing
has superseded the standard multistage matrix acidizing.
CONCLUSIONS
The following conclusions are drawn from the work performed in the Khuff reservoir:
1. For this operation, the limited entry, multiple injection
matrix acidizing technology components of stages 2 and 3
functioned successfully, as demonstrated by the pressure
signatures of launching the cutter assemblies of these stages
from the shear housings upon dropping the respective
activation balls.
2. The limited entry, multiple injection matrix acidizing
technology activation balls must be chased down the frac
string with much higher pumping rates 25 bbl/min to
35 bbl/min as compared to the normal open hole
multistage fracturing balls 5 bbl/min to 7 bbl/min.
The resultant higher fluid momentum creates a complex
ACKNOWLEDGMENTS
The authors would like to thank the management of Saudi
Aramco for their permission to publish this article. Also, special
thanks go to the stimulation team at Saudi Aramco. Furthermore, the authors would like to express a very warm and sincere
appreciation to Wael El-Mofty of Packers Plus Energy Services
for providing valuable information with regard to the technology
development and design. In addition, the authors are very
thankful for the field operational crew and their continued
dedication.
The authors would also like to extend special recognition to
Professor George V. Chilingar for his significant contributions
in advancing the knowledge about carbonate rocks.
This article was presented at the SPE Kuwait Oil and Gas
Show and Conference, Mishref, Kuwait, October 7-10, 2013.
NOMENCLATURE
max maximum horizontal in situ stress
min minimum horizontal in situ stress
kh
permeability-thickness product, md-ft
REFERENCES
1. Al-Ghazal, M.A., Al-Driweesh, S.M., Al-Ghurairi, F.A., AlSagr, A.M. and Al-Zaid, M.R.: Assessment of Multistage
Fracturing Technologies as Deployed in the Tight Gas Fields
of Saudi Arabia, IPTC paper 16440, presented at the
International Petroleum Technology Conference, Beijing,
China, March 26-28, 2013.
2. Al-Ghazal, M.A., Al-Ghurairi, F.A. and Al-Zaid, M.R.:
Overview of Open Hole Multistage Fracturing in the
Southern Area Gas Fields: Application and Outcomes,
Saudi Aramco Ghawar Gas Production Engineering
Division Internal Documentation, March 2013.
3. Al-Ghazal, M.A. and Abel, J.T.: Stimulation Technologies
in the Southern Area Gas Fields: A Step Forward in
Production Enhancement, Saudi Aramco Gas Production
Engineering Division Internal Documentation, October 2012.
4. Al-Ghazal, M.A., Al-Sagr, A.M. and Al-Driweesh, S.M.:
Evaluation of Multistage Fracturing Completion
Technologies as Deployed in the Southern Area Gas Fields
of Saudi Arabia, Saudi Aramco Journal of Technology,
Fall 2011, pp. 34-41.
5. Al-Ghazal, M.A., Al-Driweesh, S.M. and El-Mofty, W.:
Practical Aspects of Multistage Fracturing from
Geosciences and Drilling to Production: Challenges,
Solutions and Performance, SPE paper 164374, presented
at the SPE Middle East Oil and Gas Show and Exhibition,
Manama, Bahrain, March 10-13, 2013.
6. Rahim, Z., Al-Anazi, H. and Al-Kanaan, A.A.: Improved
Gas Recovery 1: Maximizing Post-Frac Gas Flow Rates
from Conventional, Tight Gas, Oil and Gas Journal, Vol.
110, No. 3, March 2012.
7. Al-Fawwaz, A., Al-Musharfi, N., Butt, P. and Fareed, A.:
Formation Evaluation While Drilling of a Complex KhuffC Carbonate Reservoir in Ghawar Field, Saudi Arabia,
SPE paper 105232, presented at the SPE Middle East Oil
and Gas Show and Conference, Manama, Bahrain, March
11-14, 2007.
8. Al-Fawwaz, A., Al-Musharfi, N., Butt, P. and Fareed, A.:
New Era of Formation Evaluation While Drilling of
Complex Reservoirs in Saudi Arabia, SPE/IADC paper
106596, presented at the SPE/IADC Middle East Drilling
and Technology Conference, Cairo, Egypt, October 22-24,
2007.
9. Ahmed, M., Rahim, Z., Al-Anazi, H., Al-Kanaan, A.A. and
Mohiuddin, M.: Development of Low Permeability
Reservoir Utilizing Multistage Fracture Completion in the
Minimum Stress Direction, SPE paper 160848, presented
at the SPE Saudi Arabia Section Technical Symposium and
Exhibition, al-Khobar, Saudi Arabia, April 8-11, 2012.
SPRING 2014
SPRING 2014
BIOGRAPHIES
Mohammed A. Al-Ghazal is a
Production Engineer at Saudi Aramco.
He is part of a team that is responsible
for gas production optimization in the
Southern Area gas reserves of Saudi
Arabia. During Mohammeds career
with Saudi Aramco, he has led and
participated in several upstream projects, including those
addressing pressure control valve optimization, cathodic
protection system performance, venturi meter calibration,
new stimulation technologies, innovative wireline
technology applications, upgrading of fracturing strategies,
petroleum computer-based applications enhancement and
safety management processes development.
In 2011, Mohammed assumed the position of Gas
Production HSE Advisor in addition to his production
engineering duties. He founded the People-Oriented HSE
culture, which has brought impressive benefits to Saudi
Arabia gas fields and resulted in improved operational
performance.
In early 2012, Mohammed went on assignment with the
Southern Area Well Completion Operations Department,
where he worked as a foreman leading a well completion
site in a remote area.
As a Production Engineer, Mohammed played a critical
role in the first successful application of several high-end
technologies in the Kingdoms gas reservoirs. Mohammeds
areas of interest include formation damage investigation
and mitigation, coiled tubing applications, wireline
operations, matrix acidizing, hydraulic fracturing and
organizational HSE performance.
In 2010, Mohammed received his B.S. degree with
honors in Petroleum Engineering from King Fahd
University of Petroleum and Minerals (KFUPM), Dhahran,
Saudi Arabia.
He has also authored and coauthored several Society of
Petroleum Engineers (SPE) papers and technical journal
articles as well as numerous in-house technical reports.
Additionally, Mohammed served as a member of the
industry and student advisory board in the Petroleum
Engineering Department of KFUPM from 2009 to 2011.
As an active SPE member, he serves on the Production
and Operations Award Committee.
Recently, he won the best presentation award at the
production engineering session of the 2013 SPE Young
Professional Technical Symposium.
Mohammed is currently pursuing an M.S. degree in
Engineering at the University of Southern California, Los
Angeles, CA.
SPRING 2014
ABSTRACT
Formation testing while drilling (FTWD) tools were introduced
as alternatives to wireline testing almost a decade ago. Interpretation of pressure tests conducted during drilling of horizontal
sections, however, is difficult because of the dynamic environment and unsteady hydrostatic pressure. One major challenge
related to pressure measurement while drilling is supercharging,
which is an increase of sandface pressure above the true reservoir
pressure. This is caused by mud filtrate invasion. The sandface
supercharge pressure can rise to greater than 1,000 psi and
cause unrealistic formation pressure estimates.
In this article, a new methodological approach has been
used to account for the effect of supercharging on formation
pressure estimations. The method begins by modeling fluid flow
within the filter cake and formation to estimate the amount of
supercharged pressure in real time. The corrected pressure is
calculated by subtracting the estimate of supercharged pressure
from the measured pressure. A new equivalent mud weight is
then calculated by using the corrected pressure. The formation
pressures obtained by the FTWD tools are taken under varying
downhole conditions to assess how forward modeling results
correlate with the analytical model results.
This new method was tested in one of Saudi Arabias fields
in real-time while drilling a horizontal section. Repeat pressure
tests were conducted a few days after the initial tests to verify
the accuracy of this mathematical model. This article discusses
the development of the supercharge pressure models, and the
results and observations from their testing.
INTRODUCTION
Wireline formation testers (WFTs) were introduced decades
ago, prompting an industry debate regarding the significance
of the tools pressure measurements. The primary questions
were how they compared to well testing results and if they
could be reliable for formation evaluation. After years of testing and continuous improvement to WFT technology, pressure
measurements from WFTs have proven to be the standard for
formation evaluation. With the introduction of formation testing while drilling (FTWD) tools, debate has arisen once more
with respect to the significance of the pressure measurements
10
SPRING 2014
of the tools. The primary concern relates to how FTWD pressure measurements compare to wireline measurements, and as
before, their reliability for formation evaluation. FTWD tools
are of interest because they can perform measurements much
sooner during the life of a well and in a potentially more
dynamic environment1.
Formation pressure measurements taken while drilling can
be affected by supercharging, which is defined as the increased pressure observed at the wellbore sandface2. Without
proper correction, supercharging can distort the pressure
readings, particularly in low permeability formations.
Pressure variations near the wellbore are primarily influenced by mud filtrate invasion and mud cake formation1. Considerable progress has been made toward understanding how
mud cakes form and influence near-wellbore pressure stability3-5.
Based on that research, miscible and immiscible multiphase
simulators have been developed to predict the filtrate invasion
for oil-based mud (OBM) and water-based mud (WBM)6. As a
result of research performed6, it is possible to make simplifying
assumptions about well conditions and still obtain a reasonable
estimate of near-wellbore pressures, although in reality the
near-wellbore is a complex environment. To control the production of formation fluids into the wellbore, wellbore pressure
is normally maintained at a pressure substantially greater than
the formation pore pressure. The wellbore sandface is exposed
to hydrostatic pressure, and the filtrate immediately invades
the near-wellbore region when a producing zone is penetrated.
Mud cake is formed when drilling fluid flows into the formation and solids are deposited on the surface of the wellbore.
This process is normally referred to as static filtration. As the
mud cake grows, it eventually stabilizes at a maximum thickness. Stabilization is a result of the shearing action of the mud
circulation in the annulus as well as the mechanical action of
the rotating drillpipe. This process is referred to as dynamic filtration. During these processes, a pressure gradient is established in the formation, Fig. 1. The pressure in the wellbore
near the surface of the mud cake is at hydrostatic (Pmh) levels,
but drops rapidly across the mud cake, and then gradually decreases across the formation, approaching formation pressure
(Pf) some distance from the wellbore. The supercharge pressure (Psc) can be defined as the difference between sandface
pressure (Pss) and Pf.
Fig. 1. Supercharging effect where hydrostatic pressure and filtration loss cause the
sandface pressure, Pss, to be elevated above the formation pressure, Pf.
The derivation of Eqn. 8 assumed that the mud cake differential, Pm, was constant, but it is not limited to this constraint. As a mud cake forms, the pressure differential changes.
In this case, the integral Pm(t)dt would simply appear in place
of Pmt. In this general form, Eqn. 8 can be used as a boundary
condition for a multiphase reservoir model where the mud
cake growth is coupled with the filtrate invasion1.
Now, by assuming the mud cake is small relative to the wellbore radius, i.e., lmc /rw 0, the following can be concluded:
(9)
(3)
Using the following non-dimensional parameters, we can
then reduce Eqn. 3 to a simpler form so that the effect of these
non-dimensional parameters can be studied.
(4)
(5)
(6)
(10)
Where:
(11)
(7)
SAUDI ARAMCO JOURNAL OF TECHNOLOGY
SPRING 2014
11
A base example was demonstrated through research1 to illustrate the supercharging effect with invasion time using the variables shown in Table 1. This example has an overbalance of
1,000 psi combined with a formation permeability of 1 md, resulting in significant dynamic supercharging effects, as demonstrated in Fig. 2. The supercharged pressure increases rapidly
Sensitivity Variable
Units
Base
Porosity (fraction)
0.25
Viscosity
(cp)
cfl (1/psi)
3 x 10-6
kmc (md)
0.0001
lmc (cm)
0.5
Wellbore Radius
rw (cm)
10
Probe Radius
rp (cm)
0.56
re (cm)
Formation Radius
rf (cm)
10,000
Formation Height
hz (cm)
10,000
Vfl (cc)
35
Vpc (cc)
Fig. 2. An example of supercharging effect using the variables from Table 1. The
dynamic mud cake growth model shows pressure increasing rapidly after exposure
to hydrostatic pressure, and as the mud cake grows, the pressure decreases. The
static model shown with the dashed line illustrates how the supercharge pressure
would increase if the mud cake were formed instantly.
12
SPRING 2014
during the very early time periods and then peaks as the mud
cake grows and chokes off the invasion. This early invasion
occurs less than a minute after the formation is exposed to the
mud hydrostatic pressure. Then the supercharge pressure
shows declination as the mud cake grows to its maximum
thickness of 0.5 cm (0.2). At this point, the pressure begins to
increase at a slow rate and approaches the dashed line showing
the results for the static mud cake model. The static mud cake
model assumes a mud cake of 0.5 cm was formed instantly
when the wellbore was exposed to hydrostatic pressure. This
research demonstrates that supercharged pressures predominately decrease when the mud cake is growing and increase at
a reduced rate when the mud cake has stabilized.
Units
Base
Porosity (fraction)
0.17
Viscosity
(cp)
cfl (1/psi)
1 x 10-5
kmc (md)
lmc (cm)
Wellbore Radius
rw (cm)
Probe Radius
rp (cm)
0.56
re (cm)
Formation Radius
rf (cm)
Formation Height
hz (cm)
Vfl (cc)
35
Vpc (cc)
(hours)
3.58
Fig. 4. The comparison of the LWD tool pressure and corrected formation
pressure after subtracting supercharge pressure during Test 2.
Fig. 3. The comparison of the LWD tool pressure and corrected formation
pressure after subtracting supercharge pressure during Test 1.
Sensitivity Variable
Units
Base
Porosity (fraction)
0.13
Viscosity
(cp)
cfl (1/psi)
1 x 10-5
kmc (md)
lmc (cm)
Wellbore Radius
rw (cm)
Probe Radius
rp (cm)
0.56
re (cm)
Formation Radius
rf (cm)
Formation Height
hz (cm)
Vfl (cc)
35
Vpc (cc)
(hours)
4.7
CONCLUSIONS
Supercharging is a result of increased sandface pressure caused
by an accumulation of mud filtrate in the wellbore region of
the formation, particularly when the permeability of a formation is low. The amount of supercharged pressure affecting the
sandface pressure measured during drilling was estimated by
using finite difference forward modeling. Corrected pressure
measurements were established by subtracting the calculated
supercharged pressure measurements from the estimated
Test
TVD (ft)
Invasion
Time (hrs)
Mobility
(md/cp)
Pstop
(psi)
Estimated
Supercharge
Pressure (psi)
Corrected
Formation
Pressure (psi)
Differences between
Repeat Test and
Corrected Formation
Pressure (psi) (Z-R)
Z1=X1-243.6
30.13
X,182.2
3.6
0.28
X1
243.6
1R
X,182.2
118
0.37
R1
R1
X,173.8
4.7
1.21
X2
91.28
Z2=X2-91.28
2R
X,173.8
87
1.22
R2
R2
86.6
Table 4. Geotap pressure tests during drilling and repeat tests while tripping out of the hole
SPRING 2014
13
formation pressure readings. Repeat tests from the same locations were performed at least 83 hours after the invasion time
of the first tests. Pressure readings showed similarity with corrected pressure readings after ignoring the supercharging effects. The most important point of this research is that new
equivalent mud weights were calculated using corrected pressures during drilling. The results proved successful for calculating new mud weight and helping accomplish optimized drilling
operations.
ACKNOWLEDGMENTS
The authors would like to thank the management of Saudi
Aramco and Halliburton for their permission to publish this
article. The authors also acknowledge the contribution of the
Saudi Aramco drilling team and Halliburton operations team
for making the field studies possible with their efforts. Special
thanks are extended to Mohammed Bayrakdar, Halliburton,
for his help with the field test and to Wael Soleiman, Halliburton, for the supercharged forward modeling described in this
research.
This article was presented at the ADIPEC 2013 Technical
Conference, Abu Dhabi, UAE, November 10-13, 2013.
REFERENCES
1. Proett, M., Chin, W.C., Lysen, S., Sands, P. and Seifert, D.:
Formation Testing in the Dynamic Drilling Environment,
paper 2004-N, presented at the SPWLA 45th Annual
Logging Symposium, Noordwijk, The Netherlands, June
6-9, 2004.
2. Wu, J., Meister, M. and Li, B.: New Method for
Supercharging Estimation, SPE paper 110389, presented
at the SPE Annual Technical Conference and Exhibition,
Anaheim, California, November 11-14, 2007.
3. Wu, J., Torres-Verdn, C., Sepehrnoori, K. and Delshad,
M.: Numerical Simulation of Mud Filtrate Invasion in
Deviated Wells, SPE paper 71739, presented at the SPE
Annual Technical Conference and Exhibition, New
Orleans, Louisiana, September 30 - October 3, 2001.
4. Chenevert, M.E. and Dewan, J.T.: A Model for Filtration
of Water-based Mud during Drilling: Determination of
Mud Cake Parameters, Petrophysics, Vol. 42, No. 3,
May-June 2001.
5. Jiao, D. and Sharma, M.M.: Mechanism of Cake Buildup
in Cross Flow Filtration of Colloidal Suspension, Journal
of Colloid and Interface Science, Vol. 162, No. 2, February
1994, pp. 454-462.
6. Chin, W.C.: Quantitative Methods in Reservoir
Engineering, Amsterdam and Boston: Gulf Professional
Publishing, USA, with Elsevier Science, July 2002, 480 p.
14
SPRING 2014
BIOGRAPHIES
Mohammed F. Al-Zayer is a
Petrophysicist in the Reservoir
Description Division of Saudi Aramco.
Since joining Saudi Aramco in 2010,
he has been involved in several
technical petrophysical disciplines of
the Ghawar field. Mohammed is
mostly interested in advanced applications of wireline and
logging while drilling formation testers as well as advanced
tools.
In 2010, he received his B.S. degree (with honors) in
Petroleum and Natural Gas Engineering from West Virginia
University, Morgantown, WV. Currently, Mohammed is
pursuing his M.S. degree in Petroleum Engineering at
Imperial College London, London, U.K.
Amer H. AbuHassoun joined Saudi
Aramco in 2001 as a Certified
Petroleum Engineer and has since then
gained hands-on experience focusing on
reservoir management, reservoir characterization, drilling and production
engineering in both sandstone and
carbonate reservoirs. He currently works in the Khurais Unit
of the Southern Reservoir Management Department as a
Senior Reservoir Management Engineer. Amers focus is on
managing the reservoir performance of the worlds largest
intelligent field: the Khurais complex. He has authored
several technical publications focusing on restoring
production utilizing an asset team approach, intelligent field
technology and enhancement of injection trends.
In 2001, Amer received his B.S. degree from King Fahd
University of Petroleum and Minerals (KFUPM), Dhahran,
Saudi Arabia, and in 2007, he received his M.S. degree
from Texas A&M University, College Station, TX, both in
Petroleum Engineering.
Dr. Sami Eyuboglu became a Program
Manager at the Halliburton Dhahran
Technology Center, Saudi Arabia, in
February 2012. He has been with
Halliburton Energy Services since April
2008. Sami specializes in both logging
while drilling and wireline pump-out
formation testers. Prior to this, he was a Research Professor
at Ohio State University, where he worked in developing
computer programs for surface geophysical methods and
numerical modeling of ground penetrating radar (GPR).
Their applications included national security issues (UXO
and tunnel detection) and the environment.
Sami received his B.S. and M.S. degrees in Mining
Engineering from Hacettepe University, Ankara, Turkey,
and his Ph.D. degree in Applied Physics from the University
of Arkansas at Little Rock, Little Rock, AR.
SPRING 2014
15
ABSTRACT
High-quality reservoir characterization, improved reservoir dynamics, optimized water flooding and a better understanding
of fluid movements are some key factors for successful reservoir management. Interwell tracer test (IWTT) technology has
been recognized as an efficient tool to determine fluid pathways between wells and evaluate areal water breakthrough
between injectors and producers, along with estimating the velocities at which the injected water is breaking through. These
data can be integrated in the geological and reservoir models
of the field to reduce uncertainties attributed to fluid flow
mechanisms and interwell communication.
This article presents a case study of IWTT technology application in a heterogeneous reservoir in a Saudi Arabian oil field.
It shows how this characterization tool was utilized to investigate the reservoir flow mechanism and how the derived information facilitated better reservoir management through improved reservoir monitoring and enhanced understanding of
reservoir fluid dynamics. The project began in November 2007
by injecting unique chemical tracers into a set of injectors to
effectively monitor injected fluid movement in the reservoir,
after which the tracers were continuously monitored through
yearly sampling programs by collecting samples from adjacent
oil producers.
The results of the project have provided valuable insights by
identifying interwell pathways, estimating velocities at which
each tracer the injected water is traveling and optimizing
water injection volumes. These findings translated into optimized reservoir management, resulting in a tangible impact on
the offset wells productivity and sustainability.
INTRODUCTION
History
SPRING 2014
Injected
Amount (kg)
Injection
Well
Well Type
Injection Date
(D-M-Y)
2-FBA
79
I-3
Horizontal
Nov. 5, 2007
4-FBA
79
I-4
Horizontal
Nov. 6, 2007
2,6-DFBA
36
I-5
Vertical
Nov. 7, 2007
3-TFMBA
13
I-1
Vertical
Nov. 3, 2007
4-TFMBA
34
I-2
Vertical
Nov. 4, 2007
18 (in P-1)
Tracer Name
METHODOLOGY
Operational Highlights
Fig. 2. Tracer breakthrough at five offset wells since the startup of tracer injection.
Frequency
SPRING 2014
17
18
SPRING 2014
Major Observations
all been proven to behave as ideal water tracers, without adsorption to the rock in either carbonate or sandstone reservoirs. The fact that two of the five tracers have been observed
in the field provides an additional confirmation of the stability
of the tracers. The chemical properties of the five tracers injected into the field are equal, so if two of them have survived
through the field, the remaining three should also survive.
Two explanations are possible for the lack of tracer production from the I-1, I-4 and I-5 injectors. One possibility, which
has also been observed in other fields, is that the injected water
and the tracers go mainly into an aquifer, where the tracers are
diluted to levels below the detection limit of 50 parts per trillion (ppt). In any tracer study design, one must consider the
possibility of dilution in water either in the reservoir or in adjacent aquifers. If a large fraction goes into the aquifer, more
than expected for a given tracer injection well, the dilution
may be too large, preventing tracer detection at the producers.
The small concentration values observed in the producers that
do produce tracers in the field (with a maximum of 10 times
the detection limit) suggest that the tracers are being diluted
more than was originally assumed in the design phase of the
project.
Another possibility is that the injected water provides pressure support to the aquifer without inducing large enough flux
within the time span of the sample collection and analysis. The
fact that the residence times are fairly long in the wells producing tracers supports the latter explanation.
Integrating a Tracers Production Curve Analysis with
Analytical Tools
tween the injectors and producers, to demonstrate flow patterns and to evaluate the sweep efficiency in this part of the
field. The curves provide significant information, such as recovered tracer mass, first tracer breakthrough and peak tracer
concentration. For instance, tracer concentration response
curves are indicative of a homogeneous reservoir if a tracer is
detected sometime after tracer injection and its detection concentration increases gradually with time. Conversely, if the
tracer concentration approaches a peak value and then decreases sharply to zero over a short period of time after the
injection start-up, it is an indication of fracture corridors or
super permeability streaks between the water injector and the
oil producer1.
As the flood front reaches an area characterized by fracture
networks, the tracer travels at a much higher velocity due to
the high-pressure differential across the fractured high permeability zone compared to that in the zone dominated with
matrix permeability. This, in turn, causes the tracer traveling
with the flood front to reach producers faster, which results in
a high tracer concentration at the offset wells. As a result, the
first tracer concentration peak can confirm the presence of a
high permeability feature connecting the injector and the
producer. An overview and analysis of the tracer response
curve for each well where tracers were detected are discussed
in detail.
Residence Time Distribution from Tracer Production Curves
SPRING 2014
19
rates, Qj, we can define the RTDs between each injector and j
as
(2)
Moment analysis of tracer curves requires that RTDs be integrated to infinity. This is not possible using measured data
alone, as any tracer campaign must be ended at some finite
time after injection. To compensate for this, integration to infinity must therefore be based on extrapolation of the tracer
curves. It has been shown4 that extrapolation of RTDs can be
done by fitting an exponential function to the tracer data for
large times. If large time data are unavailable, it may be difficult to use a log-linear fit. A different approach, based on
fitting a type curve function to the complete data set, was
therefore used6. Based on a solution to the convection-diffusion equation in known geometries9, 10, the type function with
three parameters, D0, t0 and M0, was defined and used to fit
data as:
(9)
(5)
and combined in a F diagram to quantify a measure of the
heterogeneity of the system. The swept reservoir volume as a
function of time can be estimated from F(t)5 as
(6)
Correcting Tracer Data for Reinjection
SPRING 2014
Fig. 7. RTD analysis of the 2-FBA tracers, injected at I-3 and detected in P-1, P-2, P-3, P-4 and P-5. The open circular symbols are 2-FBA data, and the dashed line is the
corresponding type curve fit defined in Eqn. 9. The light blue area corresponds to the integral of the RTD for the measured data, and the red area corresponds to the
integral of the RTD for the extrapolation. The full RTD is assumed to be the combined blue and red areas
should sum to 100% for a closed system. For a given well, j, the
recovery quantifies how much of the injected tracer is produced
in that particular well. Table 2 summarizes the recovery of the
2-FBA and 4-TFMBA tracers.
SPRING 2014
21
Fig. 8. RTD analysis of the 4-TFMBA tracers, injected at I-2 and detected in P-1 and P-5. The open circular symbols are 4-TFMBA data, and the dashed line is the
corresponding type curve fit defined in Eqn. 9. The light blue area corresponds to the integral of the RTD for the measured data, and the red area corresponds to the
integral of the RTD for the extrapolation. The full RTD is assumed to be the combined blue and red areas.
P-1
P-2
P-3
P-4
P-5
Sum
Tracer
mO [%]
mO [%]
mO [%]
mO [%]
mO [%]
mO [%]
2-FBA
0.10
0.05
0.04
0.16
0.04
0.39
4-TFMBA
0.21
0.02
0.23
Table 2. Recovery of tracers 2-FBA and 4-TFMBA, obtained from the zero moment of the RTD
Fig. 9. Tracer concentrations in wells P-1, P-2, P-3, P-4 and P-5; (a) displays responses of the 2-FBA tracer injected at I-3, and (b) displays responses of the 4-TFMBA
tracer injected at I-2.
SPRING 2014
P-1
P-2
P-3
P-4
P-5
Tracer
T [days]
T [days]
T [days]
T [days]
T [days]
2-FBA
1,662
2,099
1,296
2,369
2,185
4-TFMBA
2,046
1,564
Table 3. Average tracer residence times for the injector-producer pairs in the field
The residence times for the tracers are given by the normalized
first moment of the RTD, T = m1,j /m0,j. As these factors depend on an extrapolation to infinity, the caution for the 2-FBA
curves in P-3, P-4 and P-5, and for the 4-TFMBA curve in P-5
is considered, whereas the average residence times for the 2FBA curves in P-1 and P-2, and for the 4-TFMBA curve in P-1
are more robust.
Clearly, the average residence times, provided in Table 3, are
relatively long. The shortest one (1,296 days) corresponds to
3 years, and the longest one (2,185 days) corresponds to almost 6 years. The long residence times suggest that the injected
water moves fairly slowly through the reservoir from injector
to producer. This is indeed consistent with injection into the
aquifer and supports the observation of low recovery of tracer.
It indicates that the water injection is effective and is working
as planned.
Fig. 10. Flow capacity F vs. storage capacity based on the tracer data in P-1 and
P-2. The 2-FBA curve for P-1 characterizes the heterogeneity between I-3 (the 2FBA injector) and P-1. The 4-TFMBA curve for P-1 characterizes the
heterogeneity between I-2 (the 4-TFMBA injector) and P-1. The 2-FBA curve for
P-2 characterizes the heterogeneity between the I-3 injector and the P-2 producer.
The heterogeneity can be quantified by the normalized area between the F_ curve
and the diagonal (Lorenz coefficient).
cates the degree of heterogeneity of the reservoir. For a fractured rock, e.g., if large parts of the flow occur in a small fraction of the space, F would increase fast with increasing f .
The heterogeneity can be quantified by the Lorentz coefficient, defined by the area between the Ff curve and the diagonal, normalized by half5:
(10)
The flow capacity, F(t), and the storage capacity, f (t), were estimated from the 2-FBA tracer data in P-1 and P-2 and the 4TFMBA tracer data in P-1, using Eqn. 5. These functions are
summarized in the Ff plots in Fig. 10. Generally, the Ff
curves can be used to quantify the flow between an injector
and producer. The storage capacity, f represents the volume accessible for flow, and the flow capacity, F, represents the flow.
The curves in Fig. 10 can be used to quantify how much of the
flow occurs in a certain part of the accessible space. For example, Fig. 10 shows that about 50% of the flow occurs in about
40% of the space. This is a useful correlation to have as it indi-
SPRING 2014
23
Tracer Breakthrough In
Tracer
Injected At
P-1
I-2
4-TFMBA
2.9
18
P-1
I-3
2-FBA
10
26
P-2
I-3
2-FBA
18
P-3
I-3
2-FBA
12.2
26
P-4
I-3
2-FBA
20.1
47
P-5
I-2
4-TFMBA
13.9
44
P-5
I-3
2-FBA
22.5
46
Tracer Breakthrough In
Tracer Injected
At
Time of Breakthrough
Since Tracer Injection
Startup in Nov. 2007
(Months)
Tracer Speed
(km/month)
Flood Front
Movement
(km)
P-1
I-2
18
0.2
3.6
P-1
I-3
26
0.14
3.6
18
0.194
3.5
P-2
I-3
P-3
I-3
26
0.154
P-4
I-3
47
0.096
4.5
P-5
I-2
44
0.11
5.5
P-5
I-3
46
0.11
5.5
SPRING 2014
Tracer Breakthrough In
Tracer
Injected At
Time of Breakthrough
Since Tracer Injection
Startup in Nov. 2007
(Months)
P-1
I-2
4-TFMBA
18
35
P-1
I-3
2-FBA
26
38
P-2
I-3
2-FBA
18
38
P-3
I-3
2-FBA
26
36
P-4
I-3
2-FBA
47
44
P-5
I-2
4-TFMBA
44
49
P-5
I-3
2-FBA
46
47
Well
Name
Well Type
Tracers
Detected
From
Tracer Peak
Concentration
Time
Loss of
Circulation
Geochemical
Analysis
P-1
Horizontal
Producer
18 months (I-2)
26 months (I-3)
35 months (I-2)
38 months (I-3)
Encountered
Exhibits salinity of
injected seawater.
P-2
Horizontal
Producer
I-3
18 months
38 months
None
Exhibits salinity of
injected seawater.
P-3
Horizontal
Producer
I-3
26 months
36 months
None
Exhibits salinity of
injected seawater.
P-4
Horizontal
Producer
I-3
47 months
44 months
None
Exhibits salinity of
injected seawater
and aquifer water.
P-5
Horizontal
Producer
44 months (I-2)
46 months (I-3)
49 months (I-2)
47 months (I-3)
Encountered
Exhibits salinity of
injected seawater.
CONCLUSIONS
1. The extent of the communication between injectors and
producers has been established in one confined area along
the west flank of Field-X. The IWTT project took over a
year before registering tracer breakthrough, which is a fair
enough time given the geological complexity and
heterogeneity of Field-X.
3. Tracers injected in I-3 reached P-1, P-2, P-3, P-4 and P-5
after a period ranging from 26 to 50 months and over a
distance ranging from 3 km to 5 km.
4. The tracers injected in I-1, I-4 and I-5 have not been
detected in any of the offset wells after almost a five-year
period. This suggests either that a sort of flow barrier lies
between the three injectors and the offset wells in the north
SPRING 2014
25
Well Name
Analysis
Tracer peak concentration time indicates that the size and intensity of the fractures
connecting P-1 to I-2 have a greater adverse impact than the fractures connecting P-1 to
I-3.
P-1
Results from the image log, geoseismic attributes as shown in the Field-X Petrel model
and loss of circulation data (complete mud loss was reported) prove that major
conductive open fractures are crossing I-3. According to the image log results, the
fractures tend to be dictated in the dolomitic lithofacies, and the fracture density is
highest in the upper zone of the reservoir.
There is no proof that I-2 is intersecting any fractures since water is injected into the
matrix, rapidly flows through nearby fracture networks and leaves the matrix unswept
in this specific area.
Unlike P-1, only one tracer injected at I-3 was confirmed to show in the sample collected
from P-2. This is proof of the existence of direct communication between P-2 and I-3.
I-3 tracer arrived in at P-2 after 18 months of tracer injection, taking the same time to
arrive at the well area as that the other tracer injected in I-2 took to arrive at its offset
well, P-1.
P-2
Based on tracer detections, there is no proof that P-2 is connected to the rest of the
injectors included in this interwell tracer test.
As mentioned earlier, several pieces of evidence have confirmed that injector I-3 is
intersecting major conductive fracture networks.
Sampling was not continued on this well due to operational reasons, and as a result
tracer peak concentration was not sustainable, but took place approximately after 38
months.
This tracer took 26 months to show in the sample collected from P-3, serving as proof of
direct communication between injector I-3 and this producer.
P-3
P-4
This tracer took 36 months since its injection in I-3 in November 2007 to reach peak
concentration in P-3. This duration is more or less equivalent to the peak concentration
times of the same tracer detected in P-1 and P-2. This means the fracture networks
located between I-3 and those three producers are having nearly the same adverse
impact on those producers.
Compared to the other wells, the concentration value of the first detection was relative
ly higher. This is because the well was not being sampled prior to that. In other words,
the tracer was detected in the first sample analyzed in the lab, and there is a possibility
that the tracer could have been detected at a lower concentration prior to that if the
well was part of the sampling plan.
The concentration plot shows stable concentration values indicating that water is not
moving as fast as with the other wells included in this interwell tracer test.
The tracers injected in I-2 and I-3 were detected at higher concentrations, initially due to
the fact that this producer was not part of the sampling list until late 2011.
P-5
The tracer response curves of both tracers do not show the same shape. The concen
tration plot of the tracer injected in I-2 exhibits a bell-shaped curve; the concentration
gradually increases to a peak point in December 2011, and then gradually decreases,
forming a bell shape. This behavior indicates a strong heterogeneous system with a high
permeability channel between injector I-2 and this producer. By comparison, it is clear
that the magnitude of permeability between injector I-2 and this producer is stronger
than the magnitude between the same well and injector I-3.
Based on the sharp declining trend in tracer concentration of the I-2 tracer, if sampling
had been continued on P-5, the concentration of the tracer would have decreased
greatly until it faded.
26
SPRING 2014
NOMENCLATURE
a
b
C(t)
E(t)
F
M
Q
Qinj
Vp
(t)
ACKNOWLEDGMENTS
The authors would like to thank Saudi Aramco management
for their permission to publish this article.
This article was presented at the SPE Reservoir Characterization and Simulation Conference and Exhibition, Abu Dhabi,
UAE, September 16-18, 2013.
REFERENCES
1. Wagner, O.R.: The Use of Tracers in Diagnosing Interwell
Reservoir Heterogeneities Field Results, Journal of
Petroleum Technology, Vol. 29, No. 11, November 1977,
pp. 1,410-1,416.
2. Danckwerts, P.V.: Continuous Flow Systems, Distribution
of Residence Times, Chemical Engineering Science, Vol. 2,
No. 1, February 1953, pp. 1-13.
SPRING 2014
27
BIOGRAPHIES
Muhanad A. Al-Mosa joined Saudi
Aramco in March 2007 as a Reservoir
Engineer working in the Southern Area
Reservoir Management Department.
He has been involved in several
assignments and projects since then.
Muhanad currently works as a
Reservoir Management Engineer with part of the team
managing the giant Ghawar field. His previous experience
includes working as a Field Production Operations
Engineer, maintaining oil and water wells integrity and
productivity, and working as a Facility Engineer at the
seawater treatment and injection plants. Muhanad is
interested in reservoir and production system integration
and optimization.
He has published several technical reports, studies and
Society of Petroleum Engineers (SPE) papers, and he is a
SPE Certified Petroleum Engineer.
Muhanad received his B.S. degree in Petroleum
Engineering from King Fahd University of Petroleum and
Minerals (KFUPM), Dhahran, Saudi Arabia.
Husain A. Zaberi joined Saudi Aramco
in June 2011 as a Reservoir Engineer
working in the Southern Area
Reservoir Management Department.
Since then, he has gained a thorough
knowledge of reservoir engineering
fundamentals, substantial interpersonal
skills and up-to-date knowledge of new drilling techniques
and technologies. Husains particular area of expertise in
reservoir engineering includes coordinating with various
internal departments, such as the Reservoir Characterization and Reservoir Simulation Department, to ensure the
progress of ongoing operations and projects. Part of his
responsibilities also includes presenting reports in numerous
technical meetings and continuous monitoring of oil
production and operations.
In his current assignment as a Production Engineer,
Husain works with teams of field operators and engineers
to ensure the integrity and safety to oil wells by continuous
inspection and contingent analysis of downhole flow
control and downhole monitoring equipment.
He is a member of the Society of Petroleum Engineers
(SPE) and the Dhahran Geoscience Society (DGS).
Husain received his B.S. degree in Petroleum
Engineering from the University of Tulsa, Tulsa, OK.
28
SPRING 2014
ABSTRACT
Saudi Aramco has embarked on an aggressive plan to implement an intelligent field infrastructure and workflows across
all existing and new fields. Since the inception of the first intelligent field in 2003, the technology and processes have provided obvious benefits to Saudi Aramco and offer clear
justification for a business case in favor of the intelligent field.
Even with the evidence of such obvious gains through intelligent field implementations, very few, if any attempts have been
made to quantify the value and benefits of the program implementation in dollar terms. The exercise in quantifying the
value of a given program is considered a step toward sharing
the lessons learned within the dynamic environment of the
energy sector. It also helps with understanding how new and
improved sensors and communication technologies have combined with people and processes to transform the way business
is done today.
This article discusses some of the reasons why the value determination in the intelligent field environment is such a challenge and why such an effort raises questions from all sides, as
well as how we can present more such cases from around the
world. Additionally, this article presents some convincing
cases, more relevant to the production engineering and operations environments, quantifying the benefits in dollar terms. By
making comparisons between conventional fields and fully implemented intelligent fields using cases from within the company, the article highlights the value of the technology and
processes for those unconvinced individuals who still consider
the intelligent field as somewhat of an additional load on their
already stressed work schedules, where an overwhelming
amount of data is available and where, sometimes, the most
critical data is not available instantly.
INTRODUCTION
Historically, several individual components of the intelligent
field architecture were deployed in wells during the latter part
of the last century. The first known installation of permanent
downhole gauges dates back to 1972 in West Africa. Similarly,
the first multiphase flow meter (MPFM) went on stream in
1992. The first downhole valve later an integral component
SPRING 2014
29
VALUE DETERMINATION
Each oil company has a vision and a list of strategic goals that
the intelligent field is tailored to align with. Likewise, each oil
company has its own definition of an intelligent field. This is
why it has always been somewhat of a challenge to define
what an intelligent field encompasses. The lack of uniformity
also makes it more challenging to determine the associated
added values. From the remote monitoring and decision making involved at the drilling phase, to the routine day-to-day
production monitoring and field surveillance, intelligent field
architecture is present in one form or another at every stage of
the oil field operations. If we try to quantify the economics and
resulting benefits derived from such an implementation, spread
as it is across a wide spectrum of oil and gas operations, it
becomes quite a complex job. In fact, the project becomes so
large, involving numerous departments and asset teams, that it
can seem almost impossible to define the value chain.
It has been observed that when someone conducts a top-down,
full program valuation, which takes into consideration the entire
setup and where the value is presented in terms of increased
production, it usually produces a number that seems unrealistically high and is rarely believed, resulting in a loss of credibility.
On the other hand, trying to decouple each investment, particularly within the foundational components of IT and field automation, is a challenge, as engineers may argue over which
components provide the most benefit to the solution. For instance, a pressure sensor located at the wellhead can benefit
production monitoring as much as it can aid the monitoring of
well integrity. To have some meaning, the measurement of success has to be more specific, limited, well defined and relevant.
After a look at the above two extreme scenarios in our efforts to determine the value of the intelligent field, a middle
path was considered suitable and justifiable for the project. In
our opinion, focusing on a specific business benefit, i.e., reliability improvement, cost reduction, health, safety and environment (HSE) benefits, or similar advantages, with all the
components needed for that benefit included, will produce the
clearest value proposition to objective decision makers. Focusing on the business benefits in an intelligent field environment
and estimating the total cost of achieving the same benefits in a
conventional nonintelligent field environment will provide a
useful measure of comparison between the two cases. Yet the
results, although very specific, can still be a can of worms if
people so decide and can start a never-ending discussion. So, to
highlight the comparison, we have estimated the value of each
of the business benefits derived from the specific advantages of
the intelligent field by assuming how much it would cost to
achieve the same benefit in a conventional field environment.
In our estimates, we used existing salaries of Saudi Aramco operations personnel and equipment costs, if used at all, since using third party or vendor personnel would have elevated our
costs, increasing the measure of the value from the intelligent
field implementation.
30
SPRING 2014
FIELD SETUP
To determine the potential economic impact of the intelligent
field implementation on a project, the base state of a sample
field was compared to the intelligent field alternative for the
same field. Our intelligent field consists of three main bodies,
Khurais, Abu Jifan and Mazalij, as outlined in Fig. 1. The
main field is Khurais, which has two oil-bearing reservoirs
with an elongated north-south trending, an asymmetrical anticline structure and a lower limit with a tight aquifer2. Both
reservoirs consist of carbonate formations that are a few hundred feet thick. The upper reservoir of Khurais field also appears in fields Abu Jifan and Mazalij, which are significantly
lower producers compared to Khurais.
The north-south spread of the three main bodies is almost
200 km, whereas the east-west spread varies from 5 km to 25
km with an average of 18 km. Based on this spread, we are
looking at monitoring an area of approximately 3,600 km2.
Oil production from oil wells in the three fields is remotely
controlled and monitored using state-of-the-art sensors and
communication technologies. Each producing well is equipped
with an electric submersible pump (ESP), mounted downhole
on a Y-block to permit intervention below the ESP as and
when required. Each ESP is equipped with sensors that monitor intake and discharge pressures, motor temperature, motor
current and vibration; these parameters are critical to ensure a
smooth and efficient operation of the downhole pumps. At the
surface, a remotely controlled and monitored variable speed
drive (VSD) permits instantaneous frequency control of the
downhole ESP. Several electrical parameters like current, voltage, operating frequency, etc., are also recorded at the surface
VSD. The ESP installations play a critical role in the overall
development strategy. Often used as a tool for artificial lift, in
this case the ESP has been utilized to generate sufficient pressure at the surface to push the produced oil all the way to the
SPRING 2014
31
32
SPRING 2014
for a change in production is forwarded to the production engineering personnel, clear guidelines are sent to the operations
personnel, who affect the required changes in production almost immediately using the SCADA interfaces. Remote control
and monitoring has made life very simple, and as a result, to a
newcomer, the value realization is not there; it is construed as
business as usual.
If the same job were performed in the conventional environment, without an intelligent field setup, as an estimate it would
cost several million dollars annually just to manage the
changes in production, with field operators running from well
to well to affect the choke changes or to open and close the
wells. Additionally, since all data would have to be collected
manually, the accuracy of the collected information would always be open to question. The timeline to collect useful data,
convert it into usable information and implement a required
change was much longer in the past.
Downtime Reduction
It is a major challenge to monitor the large number of instruments installed in the wells in all of the three fields under
study. It is almost unimaginable to perform data collection on
this scale manually. The widespread area combined with fully
instrumented wells would require assembling a large workforce and overcoming the attendant logistical challenges to perform surveillance in the same manner as it is being performed
currently using the intelligent field infrastructure. An ESP installation is a significant cost item for any company. Every
effort is made to increase its run life as much as possible by
avoiding failures and preventing interventions, which result in
production loss. As seen in Fig. 2, in an intelligent field, any
ESP that stops working due to any of several possible causes is
immediately detected, and a remote startup of the ESP is
attempted. If the problem is of a serious nature, a field team is
contacted to respond immediately and provide detailed troubleshooting for the problem. A remedial plan is then enacted, and
the ESP is returned to an operational state in the shortest possible time. As is evident from the real-life example, an ESP labeled
as E tripped three times during a month and was restarted
every time it tripped. Such efficiency would be unthinkable in a
COST REDUCTION
ESP Operations
Fig. 4. Common causes of shutting down the ESP from the SCADA system.
SPRING 2014
33
To maintain the production targets set by the corporate departments, the chokes on all the producers as well as on the
water injector wells were manipulated to achieve the required
production and injection targets. In addition, the chokes were
also adjusted to optimize ESP performance at the producing
wells. On average, every choke was changed twice a month in
2012. Some wells experienced more than five choke changes in
a particular month, as illustrated by the pink line in Fig. 5. All
in all, 6,100 choke changes were logged in the SCADA system
during 2012. Just the idea of performing this task manually,
without the availability of the monitoring and control functionalities, makes it seem daunting and almost impractical. The
intelligent field has changed completely the way operations are
currently performed.
Well Performance Monitoring
34
SPRING 2014
It will not be an exaggeration to claim that the biggest gain derived from the intelligent field architecture has been in the field
of HSE. The availability of real-time data at the desktop without the need to travel hundreds of kilometers out into the
desert in extremely hot weather conditions is nothing less than
a miracle. Exposure to multiple hazards, like desert driving,
Type of Survey
SBHP on Wireline
Data Available RT
PI Test
Data Available RT
Annuli Survey
Data Available RT
Data Available RT
Every Month
ACKNOWLEDGMENTS
REFERENCES
1. A CERA Multiclient Study: The Digital Oil Field of the
Future: Enabling Next Generation Reservoir Performance,
May 2003.
2. Alhuthali, A.H., Al-Ajmi, F.A., Shamrani, S.S. and Abitrabi
Ballan, A.N.: Maximizing the Value of the Intelligent
Field: Experience and Prospective, SPE paper 150116,
presented at the SPE Intelligent Energy International,
Utrecht, The Netherlands, March 27-29, 2012.
CONCLUSIONS
1. A method to quantify the value of an intelligent field
implementation was presented.
2. Several field examples were presented to highlight the value
of the intelligent fields; for comparison purposes, the cost
SAUDI ARAMCO JOURNAL OF TECHNOLOGY
SPRING 2014
35
BIOGRAPHIES
Zaki B. Husain is a Senior Production
Engineer working with Saudi Aramcos
Southern Area Production Engineering
Department. His primary focus is on
the intelligent field aspects related to
production engineering. Prior to
joining Saudi Aramco in 2009, Zaki
worked with Schlumberger for 24 years in the production
domain, focused primarily on production enhancement,
well testing, multiphase metering and real-time monitoring
and control.
He received his B.S. degree in Electrical Engineering
from the University of Engineering and Technology,
Lahore, Pakistan.
Muhammad A. Al-Hajri joined Saudi
Aramco in 2001 as a Production
Engineer. Since that time, he has
worked in several different fields
throughout Saudi Aramco.
Muhammad became a Supervisor in
2011. His experience also includes
working for one year as a Reservoir Engineer. Muhammad
is the coauthor of several Society of Petroleum Engineer
(SPE) papers.
He received his B.S. degree in Petroleum Engineering
from King Fahd University of Petroleum and Minerals
(KFUPM), Dhahran, Saudi Arabia, in 2001.
36
SPRING 2014
ABSTRACT
This article presents a rigorous methodology for identifying the
remaining reservoir opportunities to enhance sweep efficiency and
increase hydrocarbon production rate in a mature field under
waterflooding operations. Development of mature fields relies on
proper reservoir management practices to determine the amount
and location of the remaining oil. It also entails optimal placement of new wells and reentries to enhance ultimate recovery.
As of today, common approaches to the placement of new
wells are usually based on oil column maps derived from saturation and production logs, as well as the performance of the
neighboring wells in the area of interest. The current work improves on these common practices by combining both static
and dynamic variables extracted from the geological and simulation models to estimate the reservoir opportunity index (ROI),
which can effectively detect unswept oil zones not seen by simple methods. The ROI and other available information in the
geological and simulation models, such as production history
and well logs, are then integrated into one platform to facilitate
informed decisions for successful mature field development.
The initial formulation of the ROI, previously presented by
many authors in the industry, has been improved to suit reservoir geology and conditions at different time steps, incorporating changes in oil saturation and reservoir pressure effects. The
computed ROI is then normalized to allow accurate representation and comparison among various layers and areas.
The ultimate benefits of the proposed approach are realized
when mapping the ROI for different reservoir layers and advising on the location of new wells and reentries. Moreover, the
applicability and the advantages of this work can be expanded
to other areas, such as reservoir description, reservoir characterization, and reserves estimation and depletion analysis.
INTRODUCTION
The process of identifying locations for infill drilling and the
sidetracking of existing wells in a mature field is involved. It
requires multiple levels of scrutiny to correctly estimate storage
and flow capacity, which facilitates predicting the production
performance of new wells and generating future development
plans. This process is even more complex in carbonate reservoirs
METHODOLOGY
The concept of the ROI is based on Camargo (1999)1, a paper
that set mathematical expressions to combine reservoir variables to estimate reservoir quality and assist in identifying the
best reservoir spots for future well locations. The ROI was expressed in terms of oil saturation, hydrocarbon pore volume
(PV) and flow capacity (kh) as follows:
(1)
SPRING 2014
37
GEOLOGIC OVERVIEW
The main oil-bearing formation in the field was deposited in a
shallow continental shelf sea with extensive carbonate and
evaporitic deposits. It has been divided into three main layers4.
The upper reservoir layer thickness widely varies across the
field. Due to facies changes, the reservoir quality is quite variable. Calcarenites have been developed in this layer with poor
lateral continuity. Furthermore, dolomite diagenesis and minor
anhydrite plugging have tended to degrade reservoir quality in
this layer. This layer is separated from the mid-layer by an
anhydrite streak for most of the reservoir.
The reservoir mid-layer is characterized by grain supported
carbonates. Calcarenites make up more than half of the reservoir rock, with the percentage of this lithotype varying within
the layer. Calcarenites and calcarenitic limestones are present
38
SPRING 2014
RESULTS
Reservoir Quality Index (RQI)
Two estimation scenarios are considered for the ROI computations: oil column and simulation grid, Fig. 2.
Fig. 3a. Top layer in the oil column based scenario without pressure.
Fig. 3b. Top layer in the oil column based scenario with pressure.
Regardless of its patchy character, this layer shows improvement in the ROI values when pressure is involved in the calculations. The red circles in Fig. 3b highlight the best areas to
drain, taking advantage of the waterflooding pressure support.
This analysis indicates that the pressure support in the top
layer is critical to create opportunity.
Mid-layer
Top Layer
Bottom Layer
Using the oil column based approach, the ROI in the top layer
shows very limited opportunities at the initial and final time
steps when pressure is not considered in the calculations, Fig. 3a.
SPRING 2014
39
regardless of the pressure effects, Figs. 5a and 5b. In this case, the
outputs should be handled with care, since the true oil saturation
in the wellbore vicinity is being masked by the investigation depth
of the logging tools, and oil might be present at a deeper radius.
Fig. 4a. Mid-layer in the oil column based scenario without pressure.
Fig. 4b. Mid-layer in the oil column based scenario with pressure.
40
SPRING 2014
Fig. 5a. Bottom layer in the oil column based scenario without pressure.
Fig. 5b. Bottom layer in the oil column based scenario with pressure.
Fig. 6a. Top layer in the simulation grid based scenario without pressure.
Fig. 6b. Top layer in the simulation grid based scenario with pressure.
FIELD CASE
The previously mentioned approach to calculating the ROI
was used to decide on the sidetracking of Well-A, a former vertical producer that could not sustain flow and subsequently
was converted to an observation well to monitor sweep. The
well was kept as an observation well for a long period of time,
since the old evaluation of the remaining oil in this area indicated very limited opportunity. When the area was recently revisited using the approach proposed in this article, the ROI
values showed a potential opportunity, especially using the
simulation grid based approach, Fig. 9 on the right.
The ROI values using the oil column based approach indicated a very low opportunity, which is in line with the old approaches, since oil column estimation formed the bases for
those evaluations. The ROI values at the well location, however, change from 0.2 in the oil column based scenario to 0.7
in the simulation grid based scenario. The boundary values
were used to bracket the uncertainty involved in the calculations.
Eventually, the team assigned more weight to the simulation
grid based approach and decided to sidetrack the well in the
Fig. 7a. Mid-layer in the simulation grid based scenario without pressure.
Fig. 7b. Mid-layer of the simulation grid based scenario with pressure.
CONCLUSIONS
1. The ROI computed in the present work, which is an
improved version of the concept as originally developed,
incorporates RQI as well as changes in oil saturation,
production performance and pressure.
2. The followed methodology has allowed the identification
of not only the easy opportunities, but also the ones that
are not so obvious in each reservoir layer.
3. The presented ROI maps include rock and reservoir fluid
properties, which give a higher resolution and an
engineering approach to select locations for new wells and
reentries. In addition, the calculations were carried over
multiple time steps using various approaches to delineate
SAUDI ARAMCO JOURNAL OF TECHNOLOGY
SPRING 2014
41
ACKNOWLEDGMENTS
The authors would like to thank the management of Saudi
Aramco for their permission to publish this article.
REFERENCES
1. Camargo A.: Prioritizing Opportunities for New Well
Locations and Well Workovers, paper presented at the
Geoquest Schlumberger Forum, Venezuela, 1999.
2. Amaefule, J.O., Altunbay, M., Djebbar, T., Kersey, D.G.
and Keelan, D.K.: Enhanced Reservoir Description Using
Core and Log Data to Identify Hydraulic (Flow) Units and
Predict Permeability in Uncored Intervals/Wells, SPE
paper 26436, presented at the SPE Annual Technical
Conference and Exhibition, Houston, Texas, October 3-6,
1993.
Fig. 8a. Bottom layer in the simulation grid based scenario without pressure.
Fig. 8b. Bottom layer in the simulation grid based scenario with pressure.
Fig. 9. A field case example of the oil column based approach (left) vs. a simulation
grid based approach (right), Well-A.
42
SPRING 2014
BIOGRAPHIES
Alfonso Varela-Pineda is a Petroleum
Engineer in the Southern Area
Reservoir Management Department.
Before joining Saudi Aramco in 2008,
he worked in a variety of engineering
positions with Ecopetrol, Occidental
Petroleum and Chevron in Colombia;
and with Fugro-Jason in Venezuela. Alfonso has over 10
years of diversified experience in the oil and gas industry in
areas including field operations, production, special
projects and reservoir engineering.
He received his B.S. degree from the Universidad de
America, Bogot, Colombia, and his M.S. degree from the
University of Oklahoma, Norman, OK, both in Petroleum
Engineering.
SPRING 2014
43
ABSTRACT
This article discusses the early evaluation of vertical connectivity with vertical interference tests (VITs) using an advanced
multi-probe wireline formation tester (WFT) in a giant carbonate field in Saudi Arabia. The objective was to determine the
vertical permeability of low permeability layers within the
reservoir. Early understanding of such vertical connectivity is
required to ensure the optimum development of new fields
with limited dynamic data.
To conduct the evaluation, an advanced multi-probe WFT
was utilized in five key wells around the field to obtain vertical
and horizontal permeabilities. Several VITs created pressure
pulses at the dual-packer, which produced pressure responses
monitored at two observation probes in real time. The vertical
connectivity was assessed from the pressure responses to the
transient generated across the stratigraphic layers. Comprehensive interference tests were conducted across all the layers. Advanced nonlinear regression analysis techniques were utilized
for pressure transient analysis at test intervals. The results
were further confirmed with a fine gridded 3D reservoir
simulator. The integration of all vertical permeability results
obtained so far indicate a good degree of reservoir vertical
connectivity.
The VIT results were used as input in the field simulation
model, improving the accuracy of the vertical permeability determination in different areas of the reservoir, which in turn
supported changes in the well placement strategy for maximized recovery.
layer for review. If the layers have sufficient pressure communication, there will be no pressure gradient difference across
the reservoir. This method of assessing vertical connectivity is
suitable for a field that has undergone enough production to
observe the depletion difference. An early understanding of the
vertical connectivity, however, is required to ensure optimum
development for new fields with limited dynamic data. Characterization of vertical heterogeneity is also important for well
performance and sweep efficiency.
To address this need, advanced formation testers equipped
with multiple probes and dual-packers can be utilized at the
early stage of field development to provide 3D (spatially r and z
plus time) dynamic data for the estimation of horizontal and
vertical permeability distributions along the wellbore1. It has been
shown that the integration of known static data (geology, cores
and open hole logs) and dynamic data from formation testers can
help to build more accurate reservoir models right after the
drilling phase2. Typical testing distances and the depths of investigation afforded by multi-probe formation testers can vary
between 10 ft and 100 ft, depending on reservoir properties.
When integrated with static data, results from such tests provide
more accurate evaluation and quantification of layer properties3.
Determination of vertical permeabilities with confidence
provides more accurate assessment of vertical connectivity.
Several vertical interference tests (VITs) from five wells located around the field were analyzed in this article for the early
evaluation of vertical connectivity.
INTRODUCTION
Vertical permeability is a key factor in optimum well placement to ensure excellent well productivity. Reservoirs with
good vertical connectivity allow horizontal wells to be placed
at the top of the reservoir without crossing the entire reservoir.
On the other hand, reservoirs with poor vertical communication require slanted wells, layer dedicated wells or multilateral
wells to achieve optimum volumetric efficiency.
Formation pressure testing that is usually conducted postproduction can indicate the degree of vertical connectivity
within the reservoir by producing the pressure profile in each
44
SPRING 2014
Fig. 1. Lithology and the porosity log from the study field.
can be configured with different modules to achieve the different objectives such as pressure profiling, fluid sampling and
identification, and conducting interval pressure transient tests
(IPTTs) for determining permeability distribution while
adapting to specific well and formation conditions. It is designed
for the VIT to create a pressure pulse at the source probe and
to measure observed pressures at the different layers1.
Horizontal and vertical permeabilities are obtained for each
layer along the wellbore using the dual-packer configuration
with observation probe(s). In carbonate reservoirs, formation
layering with permeabilities of different orders of magnitude
can affect productivity, recovery factors, and gas and water
breakthroughs, as well as vertical and horizontal sweep efficiencies. To assess this layering, the configuration of a dualpacker with one or two probes can be set at each flow unit
(layer or sublayer) for characterization of its permeabilities.
The number of vertically distributed IPTTs conducted depends
on the expected variation of the vertical permeability and the
existence of faults and fractures. If there are many layers (flow
units) with contrasting permeabilities and/or several fractured
or faulted zones, the number of tests should be substantial. If
there are very few distinct layers, only a few tests in each well
are needed.
For instance, if there are three layers within the radius of investigation, one IPTT with a configuration of a dual-packer
with two observation probes will be sufficient to provide pressure measurements for a well-defined inverse problem from
which horizontal and vertical permeabilities for all layers can
be determined, provided that the other properties of the formation around the wellbore, e.g., layer porosity and thickness,
are known. It should be pointed out that if the packer and two
observation probes are located in the same layer in a threelayer formation, some permeabilities for the other two layers
cannot be determined with high confidence. If the location of
the packer and those of the two observation probes are distributed among the three layers, however, then permeabilities can
be estimated with a higher degree of certainty. In this study,
three to four tests were conducted across 10 to 12 layers. It is
also important to achieve a simultaneous match on all tests
with one completed layered model.
Typical events for a configuration of a dual-packer with two
probes are shown in Fig. 2a, with pressure and rate measurements shown in Fig. 2b1, 2. One of the challenges of the IPTT
is the detection of the pressure response at the observation
probes, especially as the permeability of the formation increases or decreases while the probe distance increases. It
should be pointed out that small pressure changes can be detected and measured by high resolution quartz gauges at the
observation probes as long as the pressure changes are considerably higher than the pressure gauge resolution. Todays
quartz pressure gauges have resolutions ranging from 0.002 to
0.01 psi1. In addition, special valves are utilized to eliminate
noise, obtaining higher confidence in the case of low amplitudes at the observation probes.
SAUDI ARAMCO JOURNAL OF TECHNOLOGY
SPRING 2014
45
Fig. 2. Schematic of a WFT multi-probe packer module with two observation probes
(a) and typical pressure responses (b).
SPRING 2014
Fig. 3. A typical workflow for the interpretation of the IPTT in a layered reservoir.
incorrect layering; therefore, the geological model and its layering are also iterated.
For the estimation of the reservoir parameters, the nonlinear
estimation method is based4 on the analytical multilayer
model5, although the analytical modeling with nonlinear
parameter estimation methodology can obtain horizontal and
vertical permeabilities for a single IPTT in a formation with a
few layers. On the other hand, it can be quite challenging to
include the entire formation with many layers (10-12 in this
case) and simultaneously match pressure and derivative data
from all the IPTTs (3-4 in this case); i.e., ensure that the final
horizontal and vertical permeability estimates and the layer
definition honor all the IPTTs at the same time. The integration of all static and dynamic data all logs, images, cores
and pretests with the geological input is very important to
obtain estimates with high confidence. Although an example of
a single well was shown2, here it is demonstrated that honoring the derivative response provides significant information regarding the permeabilities of the adjacent layers in multi-well
examples. For instance, if a permeable layer (not very thick) is
surrounded with relatively low permeability layers, the derivative exhibits a radial flow regime that yields a direct estimate
of the permeability of the layer. Some of the VITs are also
matched with a fine gridded numerical model.
Field Examples
The VITs were conducted in five wells around the field for
more representative characterization. The VIT examples are
presented here from two wells due to the limited length of this
article. As discussed in the previous sections, the primary objective is to determine vertical permeability across low porosity
intervals between two geological layers.
Well-1
Fig. 5. Pressure responses and the rate during the first VIT in Well-1.
SPRING 2014
47
Fig. 10. Comparison of pressure matches measured at the packer and probes with
the layered model for the second VIT in Well-1.
Figs. 8a and 8b. Schematic of 10 VIT intervals (a) and pressure matches at the
packer and probes with the layered model for the first VIT in Well-1 (b).
H (ft)
Kh (md)
Kv (md)
1,000
100
1.5
30 - 100
30
4,000
350
3,000
350
2.5
133
13
6.5
1,000
350
280
170
1.5
2,500 - 3,000
300
1 - 10
0.1 - 1
10 - 30
1-5
Figs. 9a and 9b. Schematic of 10 VIT intervals (a) and pressure matches at the
packer and probes with the layered model for the second VIT in Well-1 (b).
SPRING 2014
objective of the work here. The increasing trend in the derivative plot after the radial flow regime at the packer is attributed
to the injection effect.
After each VIT, results were matched with the layered model
in that test interval. It was ensured that all three VIT measured
pressure responses were simultaneously matched with the same
layered reservoir model horizontal, Kh, and vertical permeabilities, Kv as a final step of the workflow. It should be
noted that the small background pressure change was subtracted for the matching process.
To increase data confidence further, the numerical model
was set up with the identical 10 layers used in the nonlinear estimation. Pressure responses were numerically simulated using
fine gridded 3D commercial software. Pressure responses at
the packer and at both probes were obtained with very close
values to the measured responses, Fig. 10.
Table 1 shows the layered model Kh and Kv results. The values
within the blue range indicate that there is a variation in permeability based on the sensitivity analysis.
Well-2
In this well, four VITs were designed and conducted with the
same dual-packer and probe distances as in Well-1. The job
objectives also included pressure measurements and sampling
at different formations in one run. The integrated plot of the
open hole wells pressure and mobility profile is shown along
with VIT locations in Fig. 11. The VIT responses and their simultaneous matching to measured pressures at the packer and
Fig. 13. Pressure match to the drawdown and buildup periods with the probe in
layer M in Well-2.
Fig. 11. VIT intervals, pressure profile and pretest mobilities in Well-2.
Kv/Kh
Figs. 12a and 12b. Schematic of VIT intervals (a) and pressure matches at the packer
and probes with the layered model for the third VIT in Well-2 (b).
Well 1
Well 2
Well 3
Well 4
Well 5
0.1
0.2
0.15
0.19
0.26
SPRING 2014
49
ACKNOWLEDGMENTS
The authors would like to thank Saudi Aramco management
for permission to publish this article.
This article was presented at the ADIPEC 2013 Technical
Conference, Abu Dhabi, UAE, November 10-13, 2013.
Fig. 14. Comparison of the VIT results and field model Kv in Well-1.
NOMENCLATURE
Table 2 summarizes the Kv /Kh results in five wells, indicating reasonably good connectivity.
Vertical permeabilities obtained from the VITs were compared with field modeling values currently in use based on
open hole logs calibrated with the cores. The comparison for
Well-1 indicated that trends are similar, Fig. 14. The other two
wells also indicated similar trends along with discrepancies in
some intervals, leading to the conclusion that field modeling
values used a constant Kv /Kh. Fine-tuning the field model with
the VIT results would increase the accuracy of the assessment
of vertical connectivity and enhance horizontal well placement.
Kh
Kv
p
CONCLUSIONS
The evaluation of vertical connectivity is the key to minimizing
the uncertainty of optimum horizontal well placement. A new
methodology to assess such connectivity was demonstrated
with comprehensive field examples, designed as a part of the
monitoring program in a layered giant carbonate reservoir. Extensive VITs conducted with special tool configurations and
best practices in well testing yielded confident measurements
despite small pressure amplitudes at the observation probes.
The flow regime analyses and nonlinear history matching
procedure, integrated with well logs and geological data, resulted in an excellent approach for estimating horizontal and
vertical permeabilities, verified by having the VITs benchmarked with a fine gridded numerical simulator. Results
showed that Kv /Kh consistently varied between 0.1 and 0.26
for the low porosity layer, indicating considerable connectivity.
Layered model results confirmed and quantified that the upper
layers (L2) have higher horizontal permeabilities up to a few
Darcy, which is a key indication for optimal and efficient horizontal well placement in these layers.
50
SPRING 2014
horizontal permeability, md
vertical permeability, md
pressure drop
REFERENCES
1. Ayan, C., Hafez, H., Hurst, S., Kuchuk, F., OCallaghan,
A., Peffer, J., et al.: Characterizing Permeability with
Formation Testers, Oilfield Review, Vol. 13, No. 3,
October 1, 2001, pp. 2-23.
2. Ma, S.M., Zeybek, M. and Kuchuk, F.J.: Integration of
Static and Dynamic Data for Enhanced Reservoir
Characterization, Geological Modeling, and Well
Performance Studies, SPE paper 166492, presented at the
SPE Annual Technical Conference and Exhibition, New
Orleans, Louisiana, September 30 - October 2, 2013.
3. Zhan, L., Kuchuk, F.J., Ma, S.M., Al-Shahri, A.M.,
Ramakrishnan, T.S., Altundas, Y.B., et al.:
Characterization of Reservoir Heterogeneity through
Fluid Movement Monitoring with Deep Electromagnetic
and Pressure Measurements, SPE paper 116328, presented
at the SPE Annual Technical Conference and Exhibition,
Denver, Colorado, September 21-24, 2008.
4. Onur, M. and Kuchuk, F.J.: Nonlinear Regression
Analysis of Well Test Pressure Data with Uncertain
Variance, SPE paper 62918, presented at the SPE Annual
Technical Conference and Exhibition, Dallas, Texas,
October 1-4, 2000.
5. Kuchuk, F.J.: Pressure Behavior of the MDT Packer
Module and DST in Crossflow Multilayer Reservoirs,
Journal of Petroleum Science and Engineering, Vol. 11,
No. 2, June 1994, pp. 123-135.
BIOGRAPHIES
Mabkhout A. Al-Harthi is a Reservoir
Engineer in Saudi Aramcos Northern
Area Reservoir Management
Department. He is currently involved
in the development of Manifa field as
the champion for the master strategic
reservoir surveillance plan. Mabkhout
has worked in a variety of disciplines, including production
engineering, drilling engineering and reservoir management.
He was involved in the reservoir management of Ghawar
field, where he conducted reservoir assessment studies,
including horizontal well performance evaluation.
Mabkhout received his B.S. degree in Petroleum
Engineering from King Fahd University of Petroleum and
Minerals (KFUPM), Dhahran, Saudi Arabia, and is
currently pursuing a M.S. degree in Petroleum Engineering,
also at KFUPM. He is certified as a Petroleum Engineer by
the Society of Petroleum Engineers (SPE) and is a SPE
member.
Cesar H. Pardo has 27 years of
experience with E&P companies. He
joined Saudi Aramco in 2006 and
worked for 1 year for the Gas
Reservoir Management Department as
a Senior Reservoir Engineer. In April
2007, Cesar was moved to the Manifa
Reservoir Management Division, where he currently works
as a Petroleum Engineer Specialist for the Manifa
Increment. He generated the Manifa reservoirs FDP and is
currently leading the Manifa Modeling and Simulation
multidisciplinary team.
In 1987, he began working at Ecopetrol (the Colombian
state oil company), where he spent 4 years involved in
drilling, workover and production technology engineering.
In 1990, Cesar joined Shell Colombia (Hocol) as a
Workover Engineer. In 1992, he was promoted to
Production Technology Engineer and successfully designed
and implemented a fracturing campaign for 30 producer
wells, and an electrical submersible pump and gas lift
campaign for over 70 wells. In 1996, Cesar was promoted
to Reservoir Engineer, working in Classical Reservoir
Engineering and Numerical Reservoir Simulation; he also
performed integrated studies to identify new infill drilling
and workover opportunities. In 2002, Cesar was promoted
to Senior Reservoir Engineer and given the additional
responsibility of Asset Exploitation Manager Acting for a
key field on production. He prepared and coordinated field
development plans to optimize sweep efficiency and
recovery in two fields with water injection. In 2004, Cesar
was promoted to Reservoir Engineering Network Leader
for the whole company in Colombia, where he coordinated
and prepared the new books with company standard
procedures to generate field forecast and calculate reserves.
Cesar received his B.S. degree in Petroleum Engineering
from the Universidad de America, Bogot, Colombia.
SPRING 2014
51
52
SPRING 2014
ABSTRACT
In reservoir surveillance, gas saturation is routinely monitored
both in gas reservoirs for reservoir performance and in saturated oil reservoirs to prevent gas coning or to optimize infill
drilling well placement. This article presents a new pulsed neutron (PN) technology and method that enables the quantitative
monitoring of the gas saturation variations to address these
reservoir management issues. One of the key features of the
newly designed PN tool is its new type of lanthanum bromide
(LaBr3) detectors, which have the best overall performance
among the detectors used in the industry. Another important new feature is its array of five detectors working together to provide an optimized solution for the targeted reservoir. The extra-long spacing of the far detectors enables a
larger volume of investigation that is more representative of
the actual reservoir condition. The quantitative aspect of the
measurement is achieved by using the ratios of the detector
counts, so that the rock matrix effects are diminished, as opposed to the traditional sigma measurement, which can be
influenced significantly by the rock matrix properties. This
new tool and data interpretation methodology have been
tested in both clastic and carbonate reservoirs with encouraging
results. This article presents an overview of the technology and
some field application examples.
INTRODUCTION
Pulse neutron (PN) technology has existed for nearly 50
years1, and during that time sigma, S , was employed as the
workhorse of saturation, S, monitoring for both oil and gas
reservoirs in high salinity environments. In 1971, Clavier et
al.2 suggested that in a quantitative saturation analysis, the difference in sigma measurements (S log) must be larger than or
equal to six cross section units (CUs), which requires:
Reservoir effective porosity (with accuracy) > 15%.
Reservoir lithology free of shales, i.e., clean formations.
Formation water salinity > 100 kppm.
The primary source of uncertainty, which results in a qualitative rather than a quantitative saturation answer outside
these conditions, is not the statistics on the measurement (S log),
but rather the uncertainty of the various input parameters
Capture Cross
Section Units
MDPN NB
Units
Pyrite
5.01
89.9
11.87
Siderite
3.94
52.3
44.69
Kaolinite
2.59
12.8
49.57
Chlorite
2.88
25.3
52.95
Illite
2.64
15.5
49.49
Quartz
2.65
4.3
19.38
Mineral
SPRING 2014
53
NEW METHODOLOGY
Recently, new measurements have emerged5 based on multi-detector pulsed neutron technology (MDPN), where measurements
sensing a larger volume of the neutron gamma transport field
are made using an array of detectors providing larger detector
source spacings than conventional instruments. These larger
sensed volumes result in higher measurement sensitivities to
several formation properties, including gas saturation6, 7.
The instrument employed for the trial described in this article comprises four spectroscopic lanthanum bromide (LaBr3)
detectors and a fast neutron detector distributed axially along
the tool body and coupled to high count rate electronics. The
instrument generates two new measurements useful in formation
evaluation: a fast neutron normalized burst (NB) and capture
ratio, derived from the nearest (proximal) and furthest (long
spaced) detectors, as well as four detector carbon/oxygen
(C/O) and sigma measurements.
Characterization of all these measurements is accomplished
using full 3D neutron-gamma transport response modeling for
the exact wellbore geometry and the borehole fluid conditions
that existed during logging of the well. The characterized instrument measurements are calibrated by a multi-point calibrator
prior to the logging; this calibrator performs a calibration of
both the magnitude and the sensitivity of the instrument readings.
Interpretation of the measurements follows mathematically
from the characterization data along traditional lines except
for the handling of shale in both the C/O and formation gas interpretations. In both cases, analytical shale characterization
and petrophysical processes have been implemented to
strengthen the mathematical definition of the shale effects and
reduce the reliance on subjective aspects of the analysis within
the saturation calculation workflow.
Previous applications of MDPN technology have been published8, 9. In this article, we address a new MDPN instrumentation
54
SPRING 2014
Fig. 1. A comparison of the new methodology (left) vs. traditional sigma (right).
with its associated nuclear attributes and its specific application to gas saturation quantification in complex mineralogical
environments.
The significant differences between this methodology and
that of traditional sigma are demonstrated in Fig. 1, and a
more detailed uncertainty analysis for the new methodology is
provided in Appendix B.
1. Traditional sigma:
Difference between sandstone oil and sandstone gas is often insufficient, making it difficult to use the technology
to differentiate gas from oil, especially if reservoir porosity
is low to intermediate.
The large effect of mineralogy, i.e., uncertainties in mineralogy determination, has a significant effect on Sw using
Eqn. 1.
2. New methodology:
The large dynamic range between sandstone oil and sand
stone gas makes the technology good for gas quantification.
Small mineralogy effect is observed, except pyrite.
The tool works in much lower porosity reservoirs.
Note: The green and red lines in Fig. 1 represent the response for sandstone oil and gas, respectively, while the other
colored lines represent the wet response for the various minerals considered; black is pyrite, blue is siderite, gray is illite,
purple is chlorite, and brown is kaolinite. In the new methodology, the gray line is exactly over the top of the brown line, so
only one is visible. The yellow bar indicates the gas sensitivity
of each system at a porosity of 20%, while the red bar illustrates the sensitivity to siderite.
ANALYSIS TECHNIQUE
En route to arriving at the saturation profile from the measured
nuclear attributes, one always goes through some important
questions. Satisfactory answers to these reinforce the validity
of the measurements and interpretation. The questions are:
Does the characterization match the formation, wellbore,
completion and borehole conditions?
Was the tool operating correctly?
Was the measurement seeing the formation?
The characterization was constructed for the exact conditions
Fig. 2. Illustration of the derivation of the in situ shale characterization; the thick
line represents the derived shale correction function from the measured data
points.
Fig. 4. Example Well-1. Track 1 displays the supplied open hole volumetrics and
correlation curves while Track 2 displays the borehole fluid condition. Track 3
displays the analysis envelope, while Tracks 4 and 5 display the PV and whole
volume results, respectively. The location of the gas-liquid contact is observed, and
an indication of which sands are developing gas caps is obtained. Remaining oil
saturation can also be estimated.
Fig. 3. Illustration of the coherence of the measured curve and the analysis
envelope. Tracks 1 and 3 contain open hole PVs and data, respectively; while
Track 2 displays the analysis envelope and measured curve. The red line of the
envelope is the gas line and the blue line is the liquid line.
Fig. 5. Example Well-2. Similar to the Fig. 4 plot, this figure shows that a gas cap
is being developed at the top of the main sand.
SPRING 2014
55
RESULTS
The MDPN NB measurement tool has been field tested in
wells completed in fluvial sandstone deposits, where pressure
CONCLUSIONS
A new measurement based on MDPN is presented with field
examples showing the applications of the MDPN technology.
The new technology is robust in reservoirs with complex
lithologies when compared to the sigma analysis technique. The
analysis technique is intuitive with a clear connection between the
derived saturations and the petrophysical inputs of shale volume
and porosity, the Monte Carlo response characterization, the
derivation of the in situ shale characterization and the generation of the analysis envelope. The trial results have been
encouraging under difficult conditions and warrant further
testing to determine the limits of the technique in terms of
porosity range, completion complexity and lithological environments. In the applications tested to date, performance of
the technique has been in line with expectations, and no
special logging practices or procedures have been required.
ACKNOWLEDGMENTS
The authors would like to thank Saudi Aramco management
for permission to publish this article.
This article was presented at the SPE Reservoir Characterization and Simulation Conference and Exhibition, Abu Dhabi,
UAE, September 16-18, 2013.
REFERENCES
1. Youmans, A.H., Hopkinson, E.C., Bergen, R.A. and Oshry,
H.I.: Neutron Lifetime, a New Nuclear Log, Journal of
Petroleum Technology, Vol. 16, No. 3, March 1964, pp.
319-328.
Fig. 6. Example Well-3. Similar to the plots in Figs. 4 and 5, this figure shows that
a gas cap is being developed at the top of the main sand.
56
SPRING 2014
Appendix A
Uncertainty Analysis of Traditional Sigma Measurement
SPRING 2014
57
Uncertainty
(V/V)
Effective
Sigma CU
Product
CU
Saturation
Uncertainty
(Sw V/V)
Total Kaolinite
Total Chlorite
Total Illite
Pyrite
0.011
67.92
0.74712
0.287354
Summation
Summation
Summation
Siderite
0.0222
37.84
0.83967
0.32295
Summation
Summation
Summation
Kaolinite
0.0315
6.24
0.19681
0.075696
Summation
Chlorite
0.0315
8.4
0.264936
0.101898
Illite
0.0315
16.24
0.51221
0.197004
Mineral
Summation
Summation
0.438861
0.444131
0.475057
Table 2. Uncertainty in gas saturation resulting from mineralogical uncertainty using sigma methodology
Uncertainty
(V/V)
Product
Ratio
Units
Saturation
Uncertainty
(Sw V/V)
Total Kaolinite
Total Chlorite
Total Illite
Pyrite
0.011
4.67
0.052
0.0027
Summation
Summation
Summation
Siderite
0.0222
10.31
0.229
0.0120
Summation
Summation
Summation
Kaolinite
0.0315
12.30
0.387
0.0204
Summation
Chlorite
0.0315
13.68
0.431
0.0227
Illite
0.0315
12.27
0.387
0.0204
Mineral
Summation
Summation
0.02393
0.02592
0.02389
Table 3. Uncertainty in gas saturation resulting from mineralogical uncertainty using the new methodology
It was prudent to go through the same error propagation exercise for this new technique as was done for sigma. To accomplish
this, full modeling was performed for the wellbore geometry to
be logged for each of the required minerals. From this, the results shown in Fig. 1 were obtained. An observation to be made
about this Monte Carlo characterization process is that ALL
neutron transport and diffusion effects are inherently accounted
for in the results. The conditions used were exactly the same as
those used for the sigma analysis, porosity of the reservoir being
0.2 V/V and with sandstone lithology as the basis. The results
of the error propagation analysis, Table 3, show remarkably
that the saturation error range is now reduced to 0.024 to
0.026, depending on which clay mineral is present an improvement factor of almost 20 times in accuracy over sigma
under these conditions. Note also the similarity in response of
the three clay types, virtually removing the need for clay type
identification and differentiation within the shale.
The picture is even clearer if the results are presented as a
ratio of the sensitivity to a particular mineral divided by the
sensitivity to gas, Table 4, where the last column is a figure of
58
SPRING 2014
Mineral
Relative Sensitivity
Merit Figure
Sigma
MDPN NB
Pyrite
26.12308
0.246568
105.9467
Siderite
14.55385
0.544351
26.73616
Kaolinite
2.4
0.649419
3.69561
Chlorite
3.230769
0.722281
4.473009
Illite
6.246154
0.647835
9.641577
BIOGRAPHIES
Mamdouh N. Al-Nasser is a Reservoir
Engineer working in Saudi Aramcos
Reservoir Description and Simulation
Department. Since joining Saudi
Aramco in 2002, he has held several
technical positions throughout the
company. Currently, Mamdouh is
working with the Petrophysical and Special Studies Unit,
where his responsibilities include reservoir saturation
surveillance technologies, integrations and best practice
optimization. His research interests are in saturation
monitoring and surveillance, nuclear logging, dynamic
petrophysics and enhanced oil recovery/improved oil
recovery surveillance design.
Mamdouh received his B.S. degree in Chemical
Engineering and M.S. degree in Petroleum Engineering in
2002 and 2012, respectively, from King Fahd University of
Petroleum and Minerals (KFUPM), Dhahran, Saudi Arabia.
He has published and coauthored several papers.
Mamdouh is a Society of Petroleum Engineers (SPE)
certified Petroleum Engineer.
Dr. Shouxiang M. (Mark) Ma is a
Senior Petrophysical Consultant in the
Reservoir Description Division and
actively serves in the Petroleum
Engineering Technologist Development
Program as a mentor and a member of
its technical review committee. He was
a founding member of the Upstream Professional
Development Center, serving as the Professional
Development Advisor for the petrophysics job family from
2009 to 2012.
Before joining Saudi Aramco in 2000, he worked as a
Lecturer at Changjiang University, Jingzhou City, China,
and as a Lab Petrophysicist at the Petroleum Recovery
Research Center in New Mexico, the Western Research
Institute in Wyoming and Exxons Production Research
Company.
Mark received his B.S. degree from the China University
of Petroleum, Beijing, China, and his M.S. and Ph.D.
degrees from the New Mexico Institute of Mining and
Technology, Socorro, NM, all in Petroleum Engineering.
He is a member of the Society of Core Analysts and the
Society of Petroleum Engineers (SPE), and he has served on
the SPEs Formation Evaluation Award Committee (as
Chairman in 2012) and the AIME/SPE Robert Earll
McConnell Award Committee.
Mark has more than 60 publications and several patents
in petrophysics. He was awarded the 2003 Department
Individual Achievement Award and 2011 SPE Saudi Arabia
Section Active Technical Involvement Award, and is a
technical journal reviewer for SPE Reservoir Evaluation
and Engineering (SPERE&E), Journal of Canadian
Petroleum Technology (JCPT), Journal of Petroleum
Science & Engineering (JPS&E) and the Arabian Journal
for Science and Engineering.
SPRING 2014
59
ABSTRACT
A practical SmartWater Flooding simulation model should
replicate laboratory and pilot-scale observations. An ideal
model should also capture the recovery mechanism in play. In
the literature, only the conventional, straightforward residual
oil reduction approach has been suggested and tested to scale
up SmartWater Flooding laboratory results. This, consequently, provides only a single view of the potential recovery
mechanism. Other approaches can be equally successful in
matching laboratory results and can provide additional insight
and understanding of the recovery mechanism in play. In this
article, we use the laboratory results of two tertiary SmartWater
corefloods to investigate the various possibilities for modeling
SmartWater Flooding and capturing the recovery mechanisms.
First, we use a high accuracy Buckley-Leverett solver to
study the performance of the two corefloods from a fractional
flow perspective. Second, we use a streamline-based simulator
to investigate in detail the possible relative permeability sets
capable of history matching SmartWater recoveries and pressures. As a result, a new set of relative permeabilities is generated
and tested through a 3D synthetic layered reservoir to demonstrate SmartWater Flooding recoveries. The results of this
work suggest that recovery enhancement through SmartWater
Flooding is best explained based on changes to the curvature of
the oil relative permeability curve, i.e., the Corey oil exponent.
Incremental recoveries are realized not due to a reduction in
the technical/absolute residual oil, but rather due to improvement in the oil flow capacity, i.e., the oil-to-water relative
permeability ratio. Based on our results, we postulate that
SmartWater incremental recovery is triggered by the formation
of a dual-wettability state across the porous medium.
INTRODUCTION
Relative permeability is the ability of a fluid phase to flow
through a porous medium in the presence of other fluid phases.
The relative permeability concept was first postulated by Muskat
and Meres (1936)1, where they extended the application of
Darcys law to two-phase systems. Relative permeabilities
are determined through either steady-state or unsteady-state
displacement experiments2. In both experiments, relative
60
SPRING 2014
EXPERIMENTAL DATA
We start our investigation with the raw coreflood data. We
correct the data to account for the apparatus dead volume.
SAUDI ARAMCO JOURNAL OF TECHNOLOGY
SPRING 2014
61
First, we correct the dimensionless time, t, by shifting the recovery (RF) to the left, such that oil production starts at zero.
Second, we also account for the dead volume by shifting the
recovery down, thereby matching the observed Sor19, Fig. 1.
Seawater Flooding
Fig. 1. Revised recovery and pressure drop data compared to the remaining oil (lines)19.
krw
kro
Sor
Swir
nw
no
Set 1
0.7
0.05
0.1441
Set 2
0.33
0.2
0.1441
2.3
2.3
Set 3
0.31
0.221
0.1441
2.5
1.8
Table 1. Corey parameters for the possible seawater relative permeability sets
Fig. 2. Fractional flow match to seawater flooding using three relative permeability sets of varying Sor (refer to Table 1); Sets 1 through 3 from left to right.
62
SPRING 2014
mal ultimate Sor of 0.05 provides the best match to the recovery profile, Fig. 2. As noted earlier, this infinitesimal ultimate Sor is in agreement with the expected performance of
oil-wet and mixed-wet systems.
(1)
At different concentrations, Cs, the modified oil relative
permeability, kro*, is calculated using Eqn. 229.
SmartWater Flooding
(2)
Using the three relative permeability sets, we investigate the independent shifts in the Corey parameters necessary to replicate
SmartWater recoveries. We assume a secondary injection of
SmartWater, i.e., we attempt to match only the ultimate recoveries
but not the production profiles. For Set 1, a reduction in the oil
exponent, no, is the only reasonable way to replicate SmartWater
recoveries, and for Sets 2 and 3, a reduction in Sor is the only possible way to match SmartWater recoveries, Table 2. Using Set 1,
Fig. 3, plots the fractional flow match against experimental data
for the 2x and 10x SmartWater injection. The oil exponent modifier a simulator input29 is also estimated to be 0.25 with
pseudo-SmartWater concentrations of 0.45 and 1 for the 2x and
10x SmartWater, respectively. As seen in Eqn. 1, the oil modifier,
mo, is related to the modified Corey oil exponent, no*.
SmartWater Effects
Figure 4 shows the associated change in the kro* due to SmartWater Flooding. Note that the increase in kro* at a given concentration as well as the shift of the intersection point toward
higher water saturations are indications of wettability modification to a more water-wet condition2, 30. This is consistent with
contact angle measurement19, Fig. 5. An approximately 20
reduction in the contact angle is reported when seawater is
substituted with the 10x SmartWater. Consistent findings have
been reported in the literature where wettability variations that
are relatively small were found to produce sizable effects on
the relative permeability curve31.
Set 1
Relative Permeability
Set 2
Set 3
Sor
Sor
no
nw
krw
0.7
0.05
0.2
0.221
2x SmartWater
3.4
0.25
0.008
0.13
0.145
10x SmartWater
300
0.0005
1,500
-0.08
0.046
0.06
Water Type
Seawater
kro
Sor
Fig. 3. Fractional flow matches to 2x (left) and 10x (right) SmartWater Flooding.
SPRING 2014
63
Fig. 6. Estimated recoveries for Set 1 (black curve) and Set 2 (red curve) against
coreflooding with seawater and SmartWater lines.
COREFLOOD SIMULATION
We start with the relative permeability for Set 1, which assumes
an ultimate Sor of 0.05, and use it in a streamline simulator to
history match the SmartWater coreflood data. The streamlinebased simulator is based on the work of Batycky et al. (1997)32
and has been extended to model low-salinity flooding29.
The Simulation Model
Fig. 5. A strong correlation between predicted changes in the oil exponent and
measured changes in the contact angle. Note that we use the average of the three
contact angle tests19.
SPRING 2014
The composite core properties are listed in Table 3. Water viscosities and the pseudo-polymer concentrations of the waters
are shown in Table 4. Using the data, we build four simulation
models with increasing complexities, Table 5. The first model
is the most simplified, assuming a constant water viscosity, a
homogeneous composite core and a constant initial water saturation of 0.1441. The second model relaxes the first assumption
in which we model the variations in the viscosities of injected
water. For this purpose, we use the simulator polymer flooding
model33 assuming that the core is initially fully saturated with a
polymer. The third and fourth models relax the homogeneity
assumption. Finally, the number of cells and their dimensions
are identical for the four simulation models. We use 118 cells
in the x-direction to minimize the error in the length of the
composite core and to reduce numerical dispersion effects.
Seawater Simulation Runs
First, we attempt to match seawater injection. Due to numerical errors across the refractive wave, the simulator overpredicts recoveries. This overprediction can be due to truncation,
roundoff and interpolation errors across the refractive wave
where the kro drops significantly. A second possible explanation
is mass destruction during the consecutive mapping between the
Eulerian and Lagrangian grids. To match both the experimental results and the analytical solution, the relative permeability
Length (cm)
Diameter (cm)
Permeability (md)
Porosity (%)
Plug 1
3.94
3.81
0.2257
74.34
Plug 2
4.16
3.81
0.2773
59.44
Plug 3
3.83
3.81
0.2497
73.26
Plug 4
3.77
3.81
0.2565
64.51
Plug 5
4.02
3.81
0.266
73.25
Plug 6
3.93
3.81
0.2036
65.26
Viscosity
(mPa.s)
Water Type
Pseudo-polymer
Concentration
Connate
0.476
Seawater
0.272
0.362
2x SmartWater
0.242
0.158
10x SmartWater
0.232
Model 1
Model 2
Model 3
Nx
118
1
0.05
0.1441
Ny
1
1
0.221
0.1441
Nz
1
1
0.2
0.1441
x (m)
0.002
y = z (m)
Model 4
3D SIMULATIONS
0.03377
Seawater
Average
Average
Varying
Heterogeneous
Heterogeneous
SPRING 2014
65
Connate
krw ratio
Seawater
2x SmartWater
mo
Cs
Sor
0.221
10x SmartWater
1.1
0.25
0.45
1
0.067
Fig. 8. Permeability porosity transforms for the top (red) and bottom (blue) zones.
low as 0.15. Second, we should keep in mind that what controls production cuts is fractional flow, the krw:kro ratio, rather
than residuals. A relative permeability curve with an Sor of 0.2
can be practically identical to another curve with an Sor of 0.
The Simulation Model
We simulate a secondary line drive waterflood. Fluid properties, i.e., oil and water viscosities as well as the initial water
saturation, are based on the previous laboratory coreflood.
The sector is 3,281 x 1,640 x 220 ft, discretized into 52 x 26 x
22 cells. The reservoir is assumed to consist of two zones, with
zone A the top 130 ft having better quality. Figure 8
shows the assumed permeability-porosity transforms along the
66
SPRING 2014
top and bottom zones. Along each zone, the reservoir is assumed to be layered with a long, normally distributed porosity,
with a 0.2 mean and a 0.25 standard deviation for zone A, and
with a 0.1 mean and a 0.2 standard deviation for zone B. Figure 9 shows the generated porosity and permeability logs. The
reservoir is produced using a line drive scheme with four producers and four injectors. The well-to-well distance is 820 ft.
The wells are completed in the top zone the top 130 ft.
Furthermore, we simulate seawater flooding and SmartWater
Flooding using two sets of relative permeability curves. The
first is the fractional flow-based curves generated in the previous section with a 0.05 Sor. The second is a more conventional
set with a 0.225 Sor to seawater flooding. Figure 10 shows the
two sets of kr curves.
Simulation Results
Secondary seawater flooding and SmartWater Flooding simulations with the two sets of relative permeability curves are performed using UT-Chem. The recovery results are plotted in
Fig. 11. Despite the significant difference in the assumed Sor
values of the two curves, the recovery predictions are almost
Fig. 10. Relative permeability sets. (Left) Fractional flow based with infinitesimal residual (0.05). (Right) Conventional curves with a practical residual definition (0.225
and 0.100 for seawater flooding and SmartWater Flooding, respectively). In both figures, seawater and SmartWater curves are shown in blue and red, respectively.
Fig. 11. UT-Chem recovery predictions against time (left), and water cut (right) for secondary seawater flooding and SmartWater Flooding using the two relative
permeability sets.
identical. The saturation profiles, Fig. 12, for secondary seawater flooding and SmartWater Flooding further demonstrate
the potential of SmartWater injection.
flow capacity, i.e., an increase in the kro:krw ratio. With SmartWater injection, the practical Sor decreases significantly due to
shifts in the relative permeability curves. With this, one might
argue that seawater flooding would recover as much as SmartWater Flooding, which could be true but it would do so
only after the injection of a impractically huge number of PVs.
SPRING 2014
67
Fig. 12. Predicted saturation profiles after 10 PVs of seawater (top) and
SmartWater Flooding (bottom), obtained using UT-Chem with the fractional flowbased relative permeability curves.
Fig. 13. Oil relative permeability curves for seawater and 10x SmartWater used in
matching laboratory corefloods. Note that the relative permeability curves are
very close to those initially predicted using fractional flow (refer to Fig. 4).
SPRING 2014
extremely low and impractical rates. As SmartWater is injected, the SmartWater front advances and alters the wettability
behind the front. Therefore, behind the front, the oil film coating
the surface starts to break and form discontinuous oil globules
in the center of the water flow paths, reducing the water capacity
to flow. Recall that a wetting phase has lower endpoint permeability than a non-wetting phase due to the essential difference
in the microscopic distribution of the remaining non-wetting
and wetting phases, respectively2. Therefore, ahead of the
CONCLUSIONS
Through analysis and simulation of unsteady-state tertiary
SmartWater displacement experiments, we have estimated the
seawater and SmartWater relative permeabilities. This information can be implemented in a commercial reservoir simulator,
e.g., Eclipse, to scale up our laboratory results. Actually, we
have investigated the various possibilities for modeling SmartWater Flooding resulting in a possible recovery mechanism.
Different relative permeability realizations have been generated
to match the experimental results. Essentially, the relative permeability realizations provide two different explanations for
the oil recovery performance in the studied carbonate system.
The first hypothesis states that no additional oil displacement recovery would take place with the continuous injection of seawater even if an infinite number of PVs are used. In
this traditional view, the SmartWater incremental is due to a
direct reduction in Sor. On the other hand, the second hypothesis
suggests that although seawater flooding has an absolute Sor of
0.05, the practical Sor is 0.221. Reaching an Sor of 0.05 would
require a huge number of PVs. With SmartWater injection, the
practical Sor decreases significantly due to shifts in the relative
permeability curves, but not the absolute residual itself.
SmartWater Flooding recovery enhancement is due to an increase in the oil flow capacity, i.e., an increase in the kro:krw ratio. Based on that, we postulate that SmartWater incremental
recovery is triggered by the formation of a dual-wettability
ACKNOWLEDGMENTS
The authors would like to thank the Saudi Aramco management and the EXPEC ARC management for granting permission to present and publish this article.
This article was presented at the SPE Reservoir Characterization and Simulation Conference and Exhibition, Abu Dhabi,
UAE, September 16-18, 2013.
NOMENCLATURE
S
C
n
m
k
kr
t
N
saturation, dimensionless
normalized concentration, dimensionless
Corey exponent, dimensionless
Corey exponent modifier, dimensionless
viscosity, ML-1T-1, Pa.s
permeability, L2, m2
relative permeability, dimensionless
dimensionless time
number of grid blocks
SUBSCRIPTS
w
o
or
wir
s
water
oil
residual oil
irreducible water
SmartWater
REFERENCES
1. Muskat, M. and Meres, M.W.: The Flow of Heterogeneous Fluids through Porous Media, Journal of Applied
Physics, Vol. 7, No. 9, September 1936, pp. 346-363.
2. Honapour, M.M., Koederitz, L.F. and Harvey, A.H.:
Relative Permeability of Petroleum Reservoirs, Boca Raton,
Florida: CRC Press, January 24, 1986, 143 p.
3. Richmond, P.C. and Watson, A.T.: Comparison of
Implicit and Explicit Methods for Interpreting
Displacement Data, SPE Reservoir Engineering, Vol. 5,
SAUDI ARAMCO JOURNAL OF TECHNOLOGY
SPRING 2014
69
16. Austad, T., Strand, S., Hognesen, E.J. and Zhang, P.:
Seawater as IOR Fluid in Fractured Chalk, SPE paper
93000, presented at the SPE International Symposium on
Oil Field Chemistry, The Woodlands, Texas, February 24, 2005.
SPRING 2014
BIOGRAPHIES
Dr. Abdulkareem M. AlSofi is a
Petroleum Engineer with Saudi
Aramcos Reservoir Engineering
Technology Team at the Exploration
and Petroleum Engineering Center
Advanced Research Center (EXPEC
ARC). His main research interest is the
modeling of enhanced oil recovery processes. Abdulkareem
has also worked with the Reservoir Management,
Reservoir Description and Simulation, and Reserves
Assessment Departments.
He is the recipient of the 2009 Society of Petroleum
Engineers (SPE) Annual Technical Conference and
Exhibition Young Professional Best Paper Award in
reservoir engineering, the recipient of the 2011 EXPEC
ARC Best Presentation Award and the recipient of the 2013
Middle East Young Engineer of the Year Award.
In 2006, Abdulkareem received his B.S. degree from the
University of Texas at Austin, Austin, TX, and in 2010, he
received his Ph.D. degree from Imperial College London,
London, U.K., both in Petroleum Engineering.
Ali A. Yousef is a Petroleum
Engineering Specialist and a Focus
Area Champion of improved oil
recovery/enhanced oil recovery
(IOR/EOR) focus area tagged
SmartWater Flood in the
Exploration and Petroleum
Engineering Center Advanced Research Center (EXPEC
ARC). He has more than 20 years of experience in
upstream research and technology. Since joining Saudi
Aramco, Ali has been involved in applied research projects
on IOR, waterflooding performance optimization, and
EOR. Currently, he is leading 40+ scientists and engineers
in developing new technologies that can increase recovery
factors from Saudi Arabian reservoirs.
Ali has written over 45 technical papers and reports and
has more than 5 patents. He is currently an active member
of the Society of Petroleum Engineers (SPE). Ali has helped
organized several petroleum engineering related conferences
and taught courses on IOR/EOR and waterflooding.
He received his M.S. and Ph.D. degrees in Petroleum
Engineering from the University of Texas at Austin, Austin,
TX.
SPRING 2014
71
72
SPRING 2014
TO ORDER
By phone/email:
SPRING 2014
73
Acknowledgments
What to send
Procedure
Author(s)/contributor(s)
Introduction
Editor
Different from the abstract in that it sets the stage for the
content of the article, rather than telling the reader what it
is about.
Main body
Conclusion/summary
74
SPRING 2014
Issue
Release date
Fall 2014
Winter 2014
Spring 2015
Summer 2015
June 1, 2014
August 31, 2014
December 1, 2014
March 1, 2015
ABSTRACT
Enhanced oil recovery (EOR) by carbon dioxide (CO2) injection is quickly becoming an important and effective method for
recovering additional oil beyond waterflooding. The CO2 EOR process is handicapped, especially in thick reservoirs, by CO2
gravity override. Due to density differences between the injected CO2 and resident fluids in the reservoir, the CO2, being lighter,
tends to rise to the top of the reservoir, thereby bypassing some of the remaining oil. Different techniques have been used to
overcome the CO2 gravity override by either increasing its density and viscosity, or by reducing its relative permeability.
Calibrating Log Derived Stress Profiles in Anisotropic Shale Gas Formations
Anas M. Al-Marzooq, Hussain A. Aljeshi and Abdullah Al-Akeely
ABSTRACT
The complex properties of unconventional gas resources pose challenges to petrophysical evaluation techniques and tools. Data
from standard logging tools and standard interpretation techniques produce high levels of uncertainties in the analysis, thereby
limiting their reliability in producing thorough petrophysical solutions. Both tight gas and shale gas formations add multiple
layers of complexity to the petrophysical evaluation, with their complex lithology and heterogeneity causing uncertainty in the
hydrocarbon volume calculations and hydraulic fracturing completion designs.
Overcoming Hydraulic Fracturing Challenges in High Temperature and Tight Gas Reservoirs of Saudi Arabia with an
Enhanced Fracturing Fluids System
Saad M. Al-Driweesh, Alaa A. Dashash, Ataur R. Malik, Jairo A. Leal Jauregui, Eduardo Soriano and Alfredo Lopez
ABSTRACT
Hydraulic fracturing has been an important aspect of the successful exploitation of gas sandstone formations in Saudi Arabia.
During the past decade, conventional formations were stimulated successfully with traditional, low to moderate temperature,
borate cross-linked fracturing fluids. As the development of the existing fields continues into deeper formations and exploration
activities are inclined toward unconventional reservoirs, new challenges are experienced due to the lower permeabilities and
higher temperatures. The conventional borate cross-linked gels are no longer the choice of fracturing fluids for extreme bottomhole conditions.
Real Life Natural Fracture Detection Examples and Workflows for Implementing Fractures in Simulation Models
Stig Lyngra, Dr. Constantine Tsingas and Nazih F. Najjar
ABSTRACT
Systematic fracture characterization is required to construct a well-constrained static and dynamic fracture model of the
reservoir. The main challenge is the need to integrate all the available data sets in a consistent manner, ranging in scale from core
samples to seismic, to allow construction of appropriate detailed geologic models and up-scaled simulation models. If this is
done with sufficient understanding of the geology and dynamic behavior of the reservoir, a history match to all available field
dynamic data can be performed. The history matched simulation model is used to generate prediction scenarios of future oil and
water production.