You are on page 1of 10

SPE 166495

Autonomous ICD Installation Success in Ecuador Heavy Oil: A Case Study


Brandon Least, Aaron Bonner, Rhandy Regulacion, Halliburton; Robert Peñaranda, Tito Sampedro, Francisco
Coloma, Repsol

Copyright 2013, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 30 September thru 2 October 2013.

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

Abstract
New technology has been called upon in Ecuador to help increase reserves and optimize the management of fluid in several
reservoirs. Autonomous Inflow Control Devices (AICD) have provided increased oil production and decreased water
production with installations in the Ginta field. When compared to neighboring wells completed with stand-alone screen
technology and producing the same total liquid, the AICD installation initially produced more oil with a lower water cut
percentage. Over several months, the water cut in the AICD installation eventually reached a similar level to that of offset
completions; however, allowed significant more oil recovery prior to reaching that point. The heavy oil in these fields is
approximately 60cP in viscosity. The wells are still in the evaluation phase; however, early data is of interest.
Due to its autonomous sensing functionality, fluid viscosity dictates the path the produced fluid will take through the
AICD. Higher viscosity oil takes a relatively non-restrictive path whereas low viscosity water is sent on a tortuous path
causing a decrease in flow rate through that particular tool. A well completed with a system of AICDs and oil swelling
packers for annular isolation restricts zones with high water ratios which in turn stimulates neighboring oil zones.
This paper presents oil production data over time for two AICD heavy oil installations in Ecuador and compares the data
to that from neighboring wells in the field. The benefit of AICD technology for reservoir management and production
optimization will be discussed.

INTRODUCTION
As reservoirs age and fields become mature, operators are turning to more complex and new techniques to optimize
remaining oil recovery. Advanced architecture wells utilize enhanced drilling techniques (Ansah et al) and multi-lateral
format to better drain oil from a field. Primary recovery is allowing the natural energy in a reservoir to drive the oil out.
When primary recovery is insufficient, operators often turn to secondary and tertiary recovery methods. Secondary recovery
is water and/or gas injection into a reservoir to maintain reservoir pressure and drive the oil out. This is referred to as water
or gas flood. Tertiary recovery includes chemicals, solvents, and heat.
Inflow control devices (ICDs) have become a common technology and are often paired with zonal isolation packers to
manage a reservoir throughout its life. They have been in use since the mid-1990s (Madsen, 1997) and by design, ICDs
create an additional choke at the completion. Traditional passive ICDs include nozzle type, tube type, and helix type and
each offers a slightly different method for creating the desired pressure drop. This intentional added pressure drop helps
offset heel-toe frictional effect within the tubing and inconsistencies in reservoir permeability. ICDs will limit flow from
becoming too high in any one zone and therefore delay water or gas from breaking through in an oil well.
Consider a full field completed with ICDs. Because the ICDs delay water and gas from breaking through, the field can
rely on primary recovery for a longer period of time. In fields where secondary recovery methods are used, the ICDs delay
the time it would take water or gas to travel from the injecting well to the producing well. Also in fields using secondary
recovery, ICDs are commonly used to balance the injection outflow in the same manner as inflow in a producing well.
Autonomous Inflow Control Devices (AICDs) are a new generation of ICDs. An AICD is an ICD that provides additional
restriction to unwanted fluids and performs this additional restriction without any connection to the surface and without any
intervention by the operator.
Two well known AICDs were described in papers SPE 159634 (Halvorsen and Elseth, 2012) and SPE 160165 (Least et
al, 2012). The later described the AICD that has moving parts while the former discussed the single phase laboratory test
2 SPE 166495

results and description of the tool function for an AICD that has no moving parts. This paper focuses on a case study in
Ecuador using the AICD that has no moving parts.
Prior to water or gas breakthrough, AICDs, like ICDs, create a choke effect to help balance influx and delay water and
gas production. Upon water or gas breakthrough, unlike ICDs, the unique autonomous ability of the AICD will change the
restriction at the breakthrough zone, creating a higher pressure differential. Figure 1 shows a theoretical comparison of a
stand-alone screen completion (SAS) vs. ICD vs. AICD completion. Figure 2 compares an ICD to an AICD completion and
shows the reduced flow rate in zone one where water has broken through. This allows more of the desired production fluid to
be produced from the other zones.

AICD restricts water, improving


the recovery slope

ICD and AICD delay


water breakthrough

Figure 1. Theoretical comparison of a stand-alone screen vs. ICD vs. AICD completions

Figure 2. Flow rate representation of water through a typical ICD and that of an AICD after breakthrough

Several key concepts about the functionality of AICDs should be noted. First, AICDs do not separate oil from water or gas.
Rather, AICDs vary the restriction according to the fluid which is passing through them. Therefore, the pressure differential
across the completion will be greater for high water and high gas zones than high oil zones. This distinctive function allows
reservoir engineers to model the completion with lower initial pressure differentials across the completion than a traditional
ICD completion.
Second, AICDs do not completely shut off the unwanted fluid, but as stated earlier, change the restriction according to the
fluid flowing through the device. A complete shutoff tool without control lines or the requirement for intervention would be
termed an autonomous inflow control valve (AICV).
SPE 166495 3

Third, if the unwanted fluid has encroached upon every zone of the well, the unwanted fluid will be produced at surface at
a high percentage. This is no surprise and may seem obvious, but misconceptions can lead to unrealistic expectations.
This paper will consider a successful case history of an AICD at the Ginta Field in Ecuador. The Ginta Field is a mature
heavy oil field, in production since 1996. Production facilities for Ginta and other fields under development by Repsol are
situated in the Amazonic jungle, which is a very delicate environment. Due to this, all the production water must be treated
and re-injected into the reservoir to minimize impact on the environment. AICDs were installed in Ginta–B100, a new well in
Ginta Field, in an attempt to increase oil production and decrease water production.
Ginta Field is a low vertical relief structure with several reservoirs. The case study is focused on a well drilled in the M1
reservoir, the most relevant reservoir regarding production in the field. M1 reservoir is a high permeability sandstone with a
very active aquifer support. The reservoir oil is 16.6° API and has a viscosity of 60cP at reservoir conditions. Due to the
highly unfavorable mobility ratio and the strong aquifer support, Ginta M1 is produced at very high water cuts (96.6% water
cut with a cumulative production of 40.6 MMbbls of oil in December 2012).
Production is achieved mainly through horizontal wells averaging 1,000ft in length at total fluid rates over 10,000bbls of
fluid per day.

Figure 3. Ginta Field Structural map, M1 reservoir. Area shaded in green represents reservoir above original oil/water contact
(OWC). Red wells are preexisting. Ginta-B100 displayed in black

COMPLETION DESIGN
The well’s horizontal (hz) production interval is a 1,000-ft long 6-1/8-in. open hole (OH) and completed with
compartmentalized screen paired with AICDs. The OH size necessitated a 4-1/2-in. lower completion string. The particle
size distribution (PSD) required a sand control screen size of 250 microns.

Constraints on Completion String Design and Estimate of Maximum Number of AICD Joints
Constraints on the completion dictated selection of completion components. Repsol’s subsurface team provided the
completion design requirement that the hz production interval should be divided into five compartments, and each
compartment with a screened sliding sleeve circulation device(SCD). Compartmentalization was achieved by utilizing four
full joint oil swellable packers. Each swellable packer has 2.3-m long element on a 16-ft basepipe, and rated to 4734 psi
pressure differential. The screened SCDs were to be used as contingency for bypassing the AICDs and opening the well to
more production if the well’s post-installation productivity turned out to be lower than expected. Figure 4 shows the
standard 4-1/2-in. SCD used in this job. Each SCD is 15.83-ft long, with 2.813 inch bore, and fitted with standard 9.25-ft
long 250 micron mesh screen jacket.
4 SPE 166495

The combined length of five SCDs and four swellable packers was approximately144 ft, which left a space of
approximately 856 ft that could be covered by maximum 21 joints of AICD-screens on range-3 basepipe. Based on this, 21
AICD joints were ordered together with the five SCDs and five swellable packers. The extra swellable packer was intended
to be used as back-up in case a section needs to be isolated.

Figure 4. 4-1/2-in. Sliding sleeve circulation device with 250 micron Mesh Screen

Design of Autonomous ICD (AICD)


The reservoir fluid conditions and expected production influenced selection of the AICD design. Oil viscosity in Ginta Field
is approximately 60cP at reservoir conditions. The applicable type of AICD in this application is Range 3, which is tested for
oil viscosities 3-200 cP. Initial pre-drilling models utilized 21 AICD Range 3A joints with each joint containing four
tungsten carbide inserts, the components of the AICD module that create resisteance to unwanted fluid flow. Using this
completion arrangement with a water breakthrough scenario, the models derived a well productivity index (PI) approximately
7 STB/d/psi. This indicated the need to increase the number of AICD inserts to accommodate the target liquid rate of up to
10,000 STB/d without too much choking of the well. The traditional way of adding AICD inserts is by increasing the
number of AICD-screen joints. However, because the length of the well could not be increased to accommodate more
equipment, the only way to increase the number of AICD inserts was by utilizing two AICD modules per joint, instead of the
typical one, thereby doubling the total number of inserts to eight per joint.
Figure 5 shows the design of each of the AICD-screen supplied for this project. Each joint is built on range-3 (38-42-ft
long) 4-1/2-in. basepipe. Only two 9.25-ft long 250 micron mesh screen jackets could be fitted, instead of the typical three,
because two AICD modules and sufficient handling spaces need to be accommodated. With each AICD module carrying up
to four inserts/valves, 168 inserts/valves were packed into 21 joints instead of the typical 84.
SPE 166495 5

Figure 5. 4-1/2-in. Autonomous ICD with 250 micron Mesh Screen

Optimization of Completion String Design


After most completion parameter s had been determined, additional modeling was done to optimize the completion. The
process of optimization of completion string design for Ginta-B-100H involved the use of petrophysical logs, primarily a
permeability profile, to determine the best possible placements of packers and distribution of AICD-screen joints to
proactively balance the inflow of fluid (Figure 6).

Figure 6. Completion Design Optimization in Using Log-Derived Permeability Profile

The first compartment (7-in. casing shoe to 9222 ft MD) has the highest average permeability, and consequently the
highest average PI per foot. Constraining this zone into a short compartment, and consequently fewer AICD-screen joints, is
desirable to give it a higher choke level relative to other compartments.
6 SPE 166495

The fourth compartment (9617-9854 ft MD) and fifth compartment (9854 ft MD to TD) have the longest intervals
between packers because they have the lowest average permeability. Their long intervals accommodated the most number of
AICD-screen joints, thereby giving them the least choke level relative to other compartments.
The second compartment (9222-9419 ft MD) and third compartment (9419-9617 ft MD) have average permeabilities
somewhere between the highest (first compartment) and the lowest (fourth and fifth compartments). Therefore, they have the
average interval lengths, average number of AICD-screen joints, and consequently average choke level.
The bottom 23 ft of the completion string consisting of float shoe, joint extension, and O-ring sub was not considered in
the pre-drilling well model. The implication is that the entire completion string would have to be adjusted up by 23 ft,
thereby affecting the optimum placements of packers. Instead of doing this, a joint of AICD-screen was removed from the
to-run tally, thereby preserving the compartmentalization design. The as-run tally, shown in Figure 7, consists of 20 AICD-
screen joints.

Figure 7. As-Run Completion Tally

METHOD OF COMPARISON
In order to quantify the additional benefit of an AICD over a stand-alone screen (SAS) completion, a method of comparison
had to be developed. A side by side well comparison can be telling, but may not be ideal since the conditions of each
reservoir are not exactly the same. Reservoir heterogeneity, different structural positions, proximity to oil-water contact
SPE 166495 7

(OWC), or drilling time differences can give misleading results. Instead of a head-to-head comparison with another well,
data from the AICD completion was used to simulate the outcome of a SAS completion then compared to actual results from
the AICD completion.
In order to best simulate the SAS completion, assumptions had to be made. All wells in the Ginta field prior to the AICD
completion had been SAS completions, so the trends observed in these wells were used to formulate the assumptions. Water
cut (WC) evolutions for the wells drilled in the 2007 drilling campaign (last campaign before the Ginta-B100) was analyzed
in addition to total fluid production over time. Well Ginta-A23 was chosen as a good comparison and was the most
influential SAS well used to model what could have been the development of water cut in a no AICDs scenario for Ginta-
B100.

Figure 8. Ginta-B100 production history

The base case used real data from Ginta-B100 completion with AICDs. Figure 8 shows the production history over time
with water cut in blue. The WC decreased from an initial high value around 83% to the well history minimum around 16%
and then slowly increased again. It was assumed that since the compartmentalization is not fully achieved due to the time
required for the swell packers to fully expand, the initial water cuts are representing the starting water cut in a non-AICD
case. With this assumption, the water evolution curve of the Ginta-A23 was truncated to the starting point measured in
Ginta-B100, and then applied to the Ginta-B100 total fluid production to determine the oil production in a non-AICD case.
Also very important to note is that Ginta-B100 was produced at a lower total fluid production rate than what is typical for
the Ginta field. The AICDs being produced at a lower total fluid production rate still produced more oil than a SAS at a
higher total fluid production rate but potentially could have been produced at a higher rate to further exploit the benefit of
AICD. This was a conclusion determined by this study.

Four scenarios were considered which are detailed in the results section.
 AICD Base Case (Real) – actual data from Ginta-B100 AICD completion
 SAS Case 1 (Simulated) – uses real total fluid production rate from Ginta-B100 but substitutes water cut from
Ginta-A23
 SAS Case 2 (Simulated) – uses higher total fluid production rate than Ginta-B100 and uses water cut from Ginta-
A23
8 SPE 166495

RESULTS AND DISCUSSION


Using the methodology described in the previous section, Figure 9 shows production data over time for both AICD
completion and what the well would have produced had it been completed with stand-alone screen (SAS). The green curve
shows actual AICD oil production over time which eventually merges with the red curve showing a SAS simulation. The
purple line shows actual AICD water cut, initially high, drops lower, then builds back up compared to the light blue line
which shows water cut for simulated SAS. The dark blue line shows actual total fluid production over time for the AICD
completion.
As noted in the introduction, once water has encroached every compartment of the completion, it is no surprise to see high
water cuts. This would be true for all completion types, AICD or other. It was believed that after the swellable packers
completed compartmentalization, that the water slowly worked its way into the other zones eventually yielding a high water
cut at surface. The benefit of the AICD was the ability to delay this from happening and recover more oil during the given
time period.

Ginta‐B‐100‐H:M1 with and without AICD's
9000 100

8000 90

80
7000

70
6000
60
5000
50
4000
40
3000
30

2000
20

1000 10

0 0
1 51 101 154 204 254 304

Ginta‐B‐100‐H:M1 BOPD real data (AICD's) Ginta‐B‐100‐H:M1 BOPD without AICD's
Ginta‐B‐100‐H:M1 BFPD  allocated Ginta‐B‐100‐H:M1 WWCT with AICD's
Ginta‐B‐100‐H:M1 WWCT without AICD's

Figure 9. Case 1 Ginta-B100 with and without AICDs

Ginta –B100 was produced at a low rate at the beginning. In a SAS with high water cuts, low production rates are not the
normal operating procedure, therefore, an additional scenario, SAS case 2, was generated in which the fluid production was
increased faster using Ginta-A23 again as a reference, with the water cut evolution defined with the previous methodology,
until it reached the actual fluid production of Ginta-B100.
Figure 10 uses the same Ginta-B100 AICD actual production data shown by the green curve (oil production), dark blue
line (total fluid production), and water cut (purple line) and compares to SAS case 2 with total fluid production shown in
orange and oil production in red. It can be seen that the total fluid production rate is much higher (orange > dark blue) which
allows the SAS case 2 oil production shown in red to be higher. The red and green oil production curves meet around 80
days instead of around day 125 in SAS case 1 but despite lower total fluid production, the AICD oil production is still higher
for this initial time period.
Figure 11 directly compares oil production over time for the actual AICD, SAS case 1, and SAS case 2 for the first 300
days of production.
SPE 166495 9

Ginta‐B‐100‐H:M1 with and without AICD's
9000 100

8000 90

80
7000

70
6000
60
5000
50
4000
40
3000
30

2000
20

1000 10

0 0
1 51 101 154 204 254 304

Ginta‐B‐100‐H:M1 BOPD real data (AICD's) Ginta‐B‐100‐H:M1 BOPD without AICD's
Ginta‐B‐100‐H:M1 BFPD without AICD's Ginta‐B‐100‐H:M1 BFPD Allocated with AICD's
Ginta‐B‐100‐H:M1 WWCT with AICD's Ginta‐B‐100‐H:M1 WWCT without AICD's
Figure 10. Case 2 Ginta-B100 with and without AICDs

BOPD
1800
1600
1400
1200
1000
800
600
400
200
0
101
111
121
131
141
151
161
171
181
191
201
211
221
231
241
251
261
271
281
291
301
1
11
21
31
41
51
61
71
81
91

Base Case (Real) Case 1 Case 2


Figure 11. Ginta-B100 cases oil production comparison
10 SPE 166495

CONCLUSIONS
 The AICD completion was able to provide benefit over other completion types. (Table 1.)
o SAS case 1 assumed same total fluid production rate as the actual data from Ginta-B100 AICD completion but
different water cut based on comparable SAS wells.
o SAS case 2 used higher total fluid production rate than Ginta-B100 and a water cut based on comparable SAS
wells.

Table 1. Recovery comparison between the presented scenarios


AICD Base Case (Real) 156 Mbbl Base Case exceeds
SAS Case 1 (Simulated) 90 Mbbl 66 Mbbl
SAS Case 2 (Simulated) 135 Mbbl 21 Mbbl

 In adition to the improvement in oil production there is also a decrease in water production that depending on the
production scenario, can be also relevant. (Table 2.)

Table 2. Total water production comparison between the presented scenarios


AICD Base Case (Real) 862 Mbbl Base Case exceeds
SAS Case 1 (Simulated) 928 Mbbl 66 Mbbl
SAS Case 2 (Simulated) 1310 Mbbl 448 Mbbl

The evolution of water cut over time in Ginta-B-100H, which showed low water cut level at initial time (first 50 days), is
the critical period for maximizing the project’s NPV (Figures 9 and 10). This period should be expoited by operating the
well at maximum rate to extract as much oil as fast as possible. This period provides a window when oil closest to the
wellbore can still outpace water before it breaks through. Operating the well at higher rate of 12,000 STB/d total liquid was
desired to attempt to extract more oil. However, the lower completion was designed for 10,000 STB/d only because of the
limitation in the number of AICDs that can be installed owing to its rather short length. The lesson learned from this first
installation that will be considered in subsequent AICD installations in Repsol’s fields in Ecuador is to design the AICD
completions to have higher initial productivity index (PI) by installing more AICDs per well. This will be achieved by
building the AICDs on 30-ft R2 or 20-ft R1 basepipes instead of the current 40-ft R3 basepipes. This will allow 33-100%
more AICD inserts to be packed into a typical 1000-ft well.

Acknowledgements
The authors wish to thank the management of Repsol, the Repsol Ecuador development team, Southwest Research Institute,
and Halliburton for permission and encouragement to present the information in this paper. We would like to thank Jason
Dykstra, Michael Fripp, and Liang Zhao for their time spent in the initial development of the Autonomous ICD product line.
We would like to thank Elibeth Mora and Byron Calvopina for support through the project.

References
Ansah, J., Shayegi, S., Ibrahim, E., “Maximizing Reservoir Potential Using Enhanced Analytical Techniques with Underbalanced
Drilling”, Paper SPE 90196, SPE Annual Technical Conference and Exhibition, 26-29 September 2004, Houston, Texas, USA
Madsen, T. “The Troll Oil Development: One Billion Barrels of Oil Reserves Created Through Advanced Well Technology”, Paper World
Petroleum Congress 29163, 15th World Petroleum Congress, October 12 - 17, 1997, Beijing, China
Halvorsen, M., and Elseth, G., “Increased oil production at Troll by autonomous inflow control with RCP valves”, Paper SPE 159634, SPE
Annual Technical Conference and Exhibition, 8-10 October 2012, San Antonio, Texas, USA
Least, B., Greci, S., Burkey, R., Ufford, A., and Wileman, A., “Autonomous ICD Single Phase Testing”, Paper SPE 160165, SPE Annual
Technical Conference and Exhibition, 8-10 October 2012, San Antonio, Texas, USA

You might also like