SPE-191961-MS

Injectivity and Well Deliverability Test and Analysis in an Offshore Complex

High CO2 Gas Carbonate Asset

Rahim Masoudi, Muhammad Afiq Abd. Wahab, Eghbal Motaei, and Abdolrahim Ataei, PETRONAS

Copyright 2018, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Asia Pacific Oil & Gas Conference and Exhibition held in Brisbane, Australia, 23–25 October 2018.

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
Among the key uncertainties and risks as part of development of a high CO2 offshore gas carbonate field;
production well deliverability, produced CO2 management, and cap rock integrity have been identified
as potential techno-commercial showstoppers that need further appraisal and studies. CO2 storage and
sequestration in the aquifer of the same field was identified as the most feasible and economic option for
the Produced CO2 management and hence the injectivity within the targeted intervals and aquifer become
part of the appraisal and study scope.
An extensive over 200 m coring program covering various intervals including overburden, caprock,
carbonate hydrocarbon intervals and aquifer has been designed as part of data acquisition and surveillance
plan. The main plan scope were designed as:

• To establish reservoir properties & characterization

• To measure formation pressure and acquire fluid sample

• To establish reservoir injectivity and productivity at the prospective intervals

• To acquire data for flow assurance analysis, facilities design and well material selection studies
for development planning.
The test and analysis has been successfully conducted covering the intended scope of the plan. Based
on the Well test and PTA, the reservoir permeability is calculated and is more or less aligned with the
core permeability with the total high skin which the majority comes from geometrical/limited-entry skin.
The productivity index is calculated to be 21 STB/day/psi. There is difficulty to analyze the Injectivity test
due to non-isothermal effects during injection and fall-off test where the fluid property of both injected
water and reservoir water is a function of temperature and time. An approximate method is applied using
the average temperature during the fall-off to simplify the case by considering a constant fluid property.
Injectivity Index is estimated from rate and pressure data to be around 26 STB/day/psi. However, it declined
by time to reach a value close to 13 STB/day/psi. In the second test, Based on pressure transient analysis
the homogeneous, vertical well with limited entry, and infinite boundary model with underneath aquifer
was accepted as representative for S2 reservoir. To capture the non-Darcy effect, the rate dependent skin
2 SPE-191961-MS

model is selected. Non-Darcy coefficient is extracted from well model for IRP in well model (1.0073E-4
(Mscf/day)-1.
Generally, the well test and injectivity and productivity analysis objectives are achieved as the fluid type
is also confirmed. The paper will detail out the actual test results, methodology and evaluation approaches
in this surveillance plan.

Introduction
The studied field is an elongated four way dip closure, platform type carbonate build up that is located in
offshore Malaysia within Sarawak Basin. The reservoir fluid is gas with high CO2 and some H2S contents.
Three well been drilled to explore and appraise the reservoir. Figure-1 depicts the reservoir cross section.
Below is the summary of the results in first two wells:
Well A1/A2
- Each well 2 DST (dedicated to infidel zones)
- Fluid Type confirmation
- Well potential is measured
- Impurities are measured
Above well results proved the viability of commercial gas production, with CO2 sequestration and
disposal in the Aquifer. However produced CO2 management and cap rock integrity become main
challenges in the field development. To narrow down the uncertainties another A3 well is drilled with the
objectives of:

• Injectivity measurement

• Evaluate Reservoir properties at Aquifer section

• Acquire core data for lab analysis and reservoir characterization

• Flow assurance Analysis

In following sections, A3 well results will be discussed to address, Well Deliverability, Injection
Test evaluation, Integrated injection & Production Analysis for reservoir characterization, Productivity
Loss estimation (vs time/PoreVolume), Fluid Type correction (Water vs CO2), Uncertainty Management,
Operational and conceptual challenges, and Evaluation of Temperature effect on Injectivity.

Figure 1—Well Tests in wells A1/A2

SPE-191961-MS 3

Well Deliverability
The well test was done in two dedicated sections including gas bearing zone and aquifer interval. First
DST was done on the aquifer interval section to test water productivity and deliverability, evaluate the
lateral heterogeneity, and to collect water sample for geochemical and petrophysical analysis. The aquifer
production test also serve to condition the wellbore prior to the injectivity test. After the flowing of the
aquifer, the test been followed with injectivity measurement in the same interval. In this section the flowing
test is discussed and in the next section will continue with injectivity analysis. Figure-2 shows the field
cross section across the field.

Figure 2—field cross section across Well A3

For well test analysis and productivity analysis, data has been assured through QA/QC and pressure gauge
reading is corrected to the mid producing interval using density profile. The flowed liquid is water and
online wellsite measurement shows gas content of the 100% CO2 as dissolved gas with H2S level of up-to
600 ppm at the maximum tested liquid rate of 4.2 kbpd. The test is a series of drawdown pulses as Multi Rate
Test (MRT). Figure-3 shows the pressure response and measured rate during short MRT periods followed
by a build which halt been decided based on the Surface Read Out (SRO) facility to optimize the testing
cost as pre-test plan was 144 hours while reduced only to 14 hours which reduced the rig time by five days.

Figure 3—History plot of Draw-Down Test of Aquifer Interval in Well A3

4 SPE-191961-MS

For the reservoir permeability estimation, build-up sections are extracted for modern derivative analysis.
Figure-4 shows derivative plot of build up periods for aquifer flow with Deconvolved pressure response.
The results confirms high skin due to partial penetration and measured permeability in this zone is about
30 mD which is lower than permeability in the upper layers from test results in wells A1/A2 which is about
45-60 mD from DST results. This lower permeability is mostly related to poorer reservoir quality in the
lower zone which aquifer lays. In order of magnitude it is in the range of core data 5 mD to 145 mD (effective
permeability to water) as depicted in Figure-5.

Figure 4—Derivative signature of two Build Up periods and Deconvolved pressure for Aquifer flow of well A3

Figure 5—DST results vs Core Data in well A3

Well model is generated using Jones for IPR model with the permeability value 33 md from PTA.
Data from span analysis is incorporated and mechanical skin is calculated using Karakas + Tariq model.
Geometrical/Partial Penetration skin is calculated by modified Cinco+Martin+Bronz model. The accuracy
of the PTA results is examined through matching the IPR/VLP model with production summary data at the
gauge level. Figure 6 shows the IPR/VLP match is good. The calculated AOF value is 101.6 Mstb/d with
the total skin of 57.14. Estimated PI with including skin factor is about 21. As compared to IPR plot, the
match quality for Darcy Vertical Well IPR gives the most closed value of AOF and total skin of 108.7 Mstb/
d and 64.6 respectively. Figure 6 illustrate the IPR together with the gauge data.
SPE-191961-MS 5

Figure 6—IPR/VLP Match for Aquifer Zone deliverability estimation

Well Deliverability of Reservoir Gas Column

This section will discuss the well test interpretation and the reservoir model based on well test data
acquired during operation for DST#2. The purpose of the analysis was to determine the reservoir properties
such as permeability, skin factor, radius of investigation, well deliverability, and the reservoir model.
During operation, there was only few gas rate measurement available at different choke size, thus for the
interpretation purposes a correlation was established based on available rate measurements to estimate the
gas rate for all drawdown intervals as illustrated in Figure 7.
These estimated rates were revised and a more accurate estimation was achieved later through the
calibrated well model where THP and BHP data at stabilized points for each chock size were used to estimate
the corresponding rate.

Figure 7—IPR/VLP Match for Aquifer Zone deliverability estimation

Mechanical and geometrical skin were estimated via the Karakas & Tariq, and Wong-Clifford models
respectively. Completion data and span analysis data were used in the model. The ratio of 0.72 between the
vertical and horizontal permeability proposed by the correlation established from core data (Figure 8) was
also used. Log interpretation indicated tow potential baffle zone within the gas column: one at the depth of
6 SPE-191961-MS

1961 to 1965 mTVD, and the other one at 2035 to 2097 mTVD (Figure 9). Thus, excluding these intervals,
a 70 m of the net pay was assumed in the model.

Figure 8—Kv/Kh correlation based on core data

Figure 9—Log for A3 well

The permeability of 355 mD resulted in an acceptable match of IPR curve with observed production data.
Calculated total skin factor is 29.25 which 17.8 out of that is ascribed to the mechanical skin and the rest
is associated to geometrical skin. A reasonable agreement between the calculated and observed production
data was achieved through the VLP matching process (Figure 10). The maximum achievable rate through
the chock size of 60/64, is estimated to be about 38 MMscf/day by the well model. It means the reported
rate of 45 MMscf/day at maximum flow period cannot be accurate. This is the tubing size that imposed
restriction on well producing rate, otherwise the well deliverability can goes much higher than that. Figure
11 shows the sensitivity analysis on tubing inside diameter.
SPE-191961-MS 7

Figure 10—VLP/IPR matching plot for DST#1

Figure 11—sensitivity analysis on tubing inside diameter

The interpretation was carried out using Well Test Analysis application. In addition to main build-up
period, the well was shut-in several times for some operational issues and also RIH/POOH for bottomhole
sampling during the test. There are five (5) build-up all together which three (3) of them offered an
acceptable quality of data to perform the analysis. This analysis was carried out based on build-up #2 (main
build-up), #1 and #5.
8 SPE-191961-MS

The vertical well with limited entry, homogeneous and infinite boundary model with an underneath
aquifer has been used to best fit the downhole data that will represent the reservoir behavior. Rate dependent
skin option was chosen and non-Darcy coefficient was considered from well model to be 1.007E-4 (MScf/
day)-1.
Figure 12 illustrates the diagnostic plot (log-log plot) along with the additional derivative of pressure
versus time for the main build-up period. Based on the curve matching, there is a small wellbore storage
effect at the beginning of the flow since well is shut-in at the wellbore, and then followed by spherical flow
due to limited entry effect of partial penetration.

Figure 12—Log-Log Derivative plot of the Build-up period for Gas Column Test

However, even before the limited entry regime is completely settled down, it is affected by a steady state
regime possibly due to existence of underneath aquifer. Scatter dp/dt points at the end tail of derivative offers
some instability in the gauge readings, however, the continued descending trend indicates the reservoir
behavior rather than wellbore dynamic effects. Data at the end of the build-up test show some influences of
the wellbore effect probably due to operational activities, but has no impact on interpretation.
Figure 13 and 14 show the superposition (semi log) plot and history matching plot respectively. The
matching quality is good and acceptable. Result from PTA confirms the same permeability value of 355
md. Total skin factor is 25.6 which is comparable with that of well model estimated.
SPE-191961-MS 9

Figure 13—Semi-Log plot of the Build-up period for Gas Column Test

Figure 14—History Match plot of the Build-up period for Gas Column Test

To evaluate the well potential, C&n model and Darcy vertical well model were applied on the main
production interval fro well modeling. The results are shown in Figure 15 and Figure 16. Regardless of
10 SPE-191961-MS

different estimation of AOF, an acceptable match with production data is achieved by the both method and
results are in agreement with that of obtained from well model.

Figure 15—IPR plot based on the Flow-After flow tests for Gas Column

Figure 16—Well Potential based on C&n Method

SPE-191961-MS 11

Pressure derivative plot is compared in Figure 17 for three (3) build-up. All are showing the same trend
and confirming the same reservoir signature. The build-up #1 and 5, however, demonstrate the longest
wellbore storage effect compared to main build-up due to shut-in the well at the surface.

Figure 17—Pressure derivative plot for different build-up periods

Figure 18 displays the permeability derived from MDT mobility points against the permeability estimated
from PTA and well deliverability modeling. Noted that mud filtrate viscosity has been assumed one (1) cp.
Both data are consistent and comparable. Thus, a conclusion can be made that rock matrix is main source
of flow capacity, and fractures (if there is any) have no contribution to the flow. But there is nevertheless
still a need for more data including core analysis data to reach a firm conclusion.
12 SPE-191961-MS

Figure 18—Permeability confirmation from different sources and Different scales

Injection Test in Aquifer

In the design stage, the feasibility of different injection fluid for the injection test was evaluated based on
the advantages, disadvantages as well as the operational risks. Based on the assessment, treated sea water
injection poses lower risk compared to liquid Nitrogen and liquid CO2 injection. The injection test was
designed based on the sensitivity done using the approved dynamic model. Poorer rock quality in the aquifer
interval has been observed in the A1/A2 wells. The expected rock quality are in the range of 10 to 40 mD
which simulated injection of between 2 kbd to 10 kbd at the maximum allowable bottomhole pressure.
Right after build up, the surface connection for injection test was prepared. Due to unavailability and
operational issues on the CO2 as injection fluid, water was injected in a step rate series with injection rate
of 1.7 kbd up to 3.7 kbd and followed by a fall off test.
During the fall off period, the well injectivity was observed to be declining. The initial skin calculated
for this Fall-off period shows higher. The bottomhole temperature reading during the injectivity shows a
reduction with the injection of the cold seawater into the bottomhole during the test. This cooling effect
to the existing fluid near the wellbore leads into the change in fluid viscosity and resulted to injectivity
reduction (Figure 19). Viscosity calculated using the Standing Correlation and bottomhole temperature over
the period of the injection were used in the Semi-steady state Darcy equation to derive the skin factor for
the well. The calculated skin factor derived using this varying viscosity are within acceptable range and
SPE-191961-MS 13

show negligible tolerance. This can be used as an evidence to support that the cooling effect resulted to the
injectivity reduction and not due to plugging or any variable mechanical skin effect.

Figure 19—Temperature cooling effect on Productivity due to viscosity change

Fall off period analysis shows higher skin with permeability of 29 mD, same as build test (prior to
injection test). This phenomenon is due to cooling effect of injected fluid in the near wellbore which leads
to productivity loss as a response of fluid viscosity to transient temperature, which it will be discussed in
detail later. Figure 20 shows a PTA analysis of fall off test. Derivative plot showing mask of radial flow
by constant pressure boundary. This is due to near gas-water contact testing. However to clarify that the
response is coming from reservoir boundary, normal derivate plot is shown in Figure 20. As shown, normal
derivative (dotted triangles) is descending all the way, indicating the response is coming from reservoir not
a wellbore induced.

Figure 20—Fall off test Derivative signature

14 SPE-191961-MS

Integrated Injection & Production Analysis for Reservoir Characterization

Individual analysis of the injection test and production tests providing different results for permeability and
also skin factor, primarily resulted from reservoir cool down. To have an integrated analysis of whole test, an
unsophisticated single well model is built in Reveal application with the physical configuration of the fluid
contacts and it is initialized using aquifer flow after flow test results and petrophysical properties of the well.
The historical data is assigned to the layer cake model in a rate control model to and reservoir pressure
requested at the gauge depth to avoid gauge shift depth. The pressure been matched in full historical cycle
(production, build up, injection and fall off) and is plotted against overall historical pressure gauge data in
Figure 21. All flow periods are having reasonable match.
At the injection test the medium rate has been measured to be constant while the pressure data does not
support that, in other words reported fixed rate for the second step rate interval is not fully trusted while
other injection test stages are accepted to be fixed and maintained in each stage.

Figure 21—Integrated Production & Injection Well Test Analysis

Productivity Loss Estimation

During production test, the Productivity Index (PI) is measured to be as high as 22 RB/Day/Psi at all
drawdown stages consistently, while during injection test it is been lost down to 13 RB/Day/Psi. Figure-8
depicts the PI loss during test progress mainly in Injectivity test period. Even though PI loss is significant
within short period of injection, still not fully captures the total loss of PI in long term. To estimate, a long
term injection, Reveal model is requested to generate the PI profile for the well matched with DST test
results, the injection been continued for years. The PI has reduced to less than 4 after few months as depicted
in Figure 22.
As a matter of fact, the productivity loss is due to limited volume that could be accepted by the reservoir
through the time. The injectivity loss happens to increase in average reservoir pressure along the time.
Figure 23 showing decrease in the injection pressure margin versus time. The limitation of injection pressure
is defined to avoid exceeding allowable injection pressure (i.e. inject below fracture gradient). PI decline
SPE-191961-MS 15

could be arrested to some extent by producing from gas zone, however due to aquifer support main PI loss
still limiting the injection rate.

Figure 22—Short Term Productivity Index Loss in well A3 matched to the historical data

Figure 23—Forecasted Long Term Productivity Index Loss in well A3

16 SPE-191961-MS

Fluid Type Correction (Water vs CO2)

As mentioned before, due to operational challenges it was decided to inject water instead of CO2. Due to
higher compressibility of CO2 compare to water, the PI is maintained at all through the time with CO2
injection as shown in Figure 24. However, CO2 injection rate is declining as reservoir pressure is increasing.
The injection profile is depicted in Figure 25.
The CO2 injectivity ideally should be higher than calculated from water injection due to heat convection
to CO2 due to compression at compressor stages and lower heat conductivity in the wellbore. As injected
fluid temperature rises, aquifer cool down decreases and will help to have better efficiency and higher
injection rate. In addition to this analysis and simulation study, the CO2 injection test results in Long core
have been conducted and the results will be incorporated into the simulation study when it is compared to
water injection case.
Applying 6250 psi as allowable average reservoir pressure within the layer cake model, 52 Bscf CO2
cummulatively could be injected as per a single well in the reservoir. Technical full field injectivity should
be optimized based on final well count, reservoir heterogeneity, gas production, structure geometry, and
evacuation strategy.

Figure 24—Forecasted Long Term Productivity Index Loss in well A3 corrected to CO2 injection
SPE-191961-MS 17

Figure 25—Forecasted Injection rate and average reservoir pressure profile for CO2 injection

Conclusion
• Challenges of a high CO2 field Development and Management has been addressed in this paper
when it comes to well productivity and CO2 injection/storage.
• A comprehensive data acquisition plan is designed to address various challenges and concerns
with regards to the reservoir properties & characterization, fluid analysis, rock mechanical tests,
caprock integrity, etc.
• Series of tests and analysis have been conducted to evaluate the well deliverability, productivity
behavior, injectivity analysis, etc. using gas production and water injection tests.
• As the main injectant later is CO2, a combination of CO2 core test analysis together with the
designed simulatiuon study on both water and gas injection have been conducted to evaluate the
actual production and injection behavior and performance at real field condition.
• The designed data acquisition and post analysis shown briefly in thi paper significantly helps to
address several uncertainties and concerns moving toward the development planning in this field.

Acknowledgment
The authors would like to appreciate PETRONAS for permission to have this paper published.

References
Mohd Azran Abd Jalil, Rahim Masoudi, Nasir H. Darman, Mohamad Othman, SPE, PETRONAS, "Study the CO2
Injection and Sequestration in Depleted M4 Carbonate Gas Condensate Reservoir, Malaysia", CMTC 150050, Carbon
Management Technology Conference held in Orlando, Florida, USA, 7–9 February 2012.
Mohd Azran Abd Jalil, Rahim Masoudi, Nasir H. Darman, Nor Zarith Sofia, Mohamad Othman, SPE, PETRONAS
"Challenges in Improving the Hydrocarbon Gas Recovery: Case Study Using a Carbonate Gas Condensate Field in
18 SPE-191961-MS

Malaysia", SPE 145867, SPE Asia Pacific Oil and Gas Conference and Exhibition held in Jakarta, Indonesia, 20–22
September 2011.
Rahim Masoudi, Mohd Azran Abd Jalil, PETRONAS; Chee Phuat Tan, David Press, John Keller, Leo Anis, Schlumberger;
Nasir Darman and Mohamad Othman, PETRONAS, "Simulation of Chemical Interaction of Injected CO2 and
Carbonic Acid Based on Laboratory Tests in 3D Coupled Geomechanical Modeling", IPTC 16703, International
Petroleum Technology Conference held in Beijing, China, 26–28 March 2013.
Rahim Masoudi and Mohd Azran Abd Jalil, PETRONAS; David Press, Kwang-Ho Lee, Chee Phuat Tan, and Leo
Anis, Schlumberger; Nasir Darman and Mohamad Othman, PETRONAS, "An Integrated Reservoir Simulation-
Geomechanical Study on Feasibility of CO2 Storage in M4 Carbonate Reservoir, Malaysia", IPTC 15029, International
Petroleum Technology Conference held in Bangkok, Thailand, 15–17 November 2011.