SPE 109294
A CO2-Rich Gas Well Test and Analyses
Q. Xu, SPE, and G. Weir, SPE, CO2CRC/Curtin University of Technology; L. Paterson, SPE, CO2CRC/CSIRO
Petroleum; I. Black, Upstream Petroleum; and S. Sharma, SPE, CO2CRC/Schlumberger Australia
Copyright 2007, Society of Petroleum Engineers
This paper was prepared for presentation at the 2007 SPE Asia Pacific Oil & Gas Conference
and Exhibition held in Jakarta, Indonesia, 30 October–1 November 2007.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes 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 where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.
AbstractThis paper reports on the planning, operation, results
and analysis of a carbon dioxide (CO2) well test performed on
Buttress-1, a CO2-rich gas well located in the Otway basin,
Victoria, Australia. A field-scale study of CO2 sequestration is
planned by the Australian Cooperative Research Centre for
Greenhouse Gas Technologies (CO2CRC) in this area, which
will involve, inter alia, taking CO2 from the Buttress reservoir
and injecting it into a nearby depleted gas field. Understanding
the production characteristics of the well and gas initial in
place of the field is important to the success of this project.
Testing comprised of a multiple rate test and extended
drawdown test in order to determine general well and reservoir
characteristics, such as the well deliverability, the non-Darcy
coefficient and the average reservoir permeability and volume.
The well test operations were challenging, as there have
not been many instances where CO2 wells have been flow
tested in the past. Hydrate formation, flow assurance, and
environmental concerns, such as noise control, soil
contamination, and safety level of CO2 were considered in the
operation. Operations were completed successfully with no
incidents and the required data were obtained successfully.
Compared to the wealth of experience developed over
many years in testing oil and gas wells, the collective
experience in CO2 well testing is limited. The distinguishing
features between this test and those of a typical natural gas
well test need to be emphasised. Although, in general, flow
testing a CO2 well should be similar to testing a natural gas
well, differences in the thermodynamic properties of CO2
affect the analysis of the well test considerably. In particular,
the non-Darcy effect can be more pronounced and the
wellbore and surface flow can involve dramatic phase
changes, such as the formation of ice. In addition, since CO2 is
more compressible than a typical natural gas, the accurate
measurement of the flow rate becomes more challenging. It is
also apparent that the use of pseudo pressure, as opposed to
simpler methods of dealing with the pressure dependency of
key properties, is essential to the successful analysis of the
pressure response to the CO2 production.
Introduction
The Australian Cooperative Research Centre for Greenhouse
Gas Technologies (CO2CRC) is planning to carry out a field-
scale CO2 sequestration demonstration project in the Otway
Basin, Victoria, which will involve injecting of the order of
100,000 tonnes of CO2 from the Buttress field into a nearby
depleted gas field.
The Buttress field is a fault-confined CO2 reservoir with a
single potential producing well, Buttress-1. This well was
drilled to a total depth of 1,732 metres, penetrating the CO2
bearing Waarre C formation at 1616 m, and then the Waarre A
sandstone formation, before being logged, completed, and
suspended as a potential CO2 producer.
Over the period 22-27 June 2006, at the request of the
CO2CRC, SGS Expertest and Upstream Petroleum carried out
a comprehensive well test on Buttress-1. This was a necessary
step in the finalisation of this project execution plan. In this
paper, the well test planning, procedure, results, and analysis
for the Buttress-1 CO2 well test are reported.
There are only a few subsurface CO2 accumulations that
have been explored for by industry with a view to commercial
production and this means that, compared with the wealth of
experience developed over many years in testing oil and gas
wells, the collective experience in CO2 well testing is
relatively modest.
In principle, testing a CO2-rich well, typically with a CO2
content in excess of 60 mol%, should not be markedly
different from a standard well test for a natural gas well.
However, the thermodynamic properties of CO2 result in
distinctly different flow behaviour in all parts of the
production system (subsurface, wellbore, and surface
facilities) making the design and reliable analysis of well test
data challenging. Jokhio et al. (2001) analysed both field data
and simulated examples and concluded that significant errors
can result from CO2 well test interpretations if methods for
natural gas are used without properly taking into account the
CO2 properties. The error can be in excess of 75% in
2 SPE 109294
deliverability tests, and between 15 and 30% when estimating
the open flow potential.
Pure CO2 in a reservoir deeper than about 800 m is usually
above its critical pressure (1071 psi, 7.39MPa) and
temperature (31.3oC). In this state it is a dense fluid (about
350 kg/m3 in this case study) having a relatively low viscosity
(0.03cp) for a fluid. In general, at high flow rates, the
interstitial velocity near the wellbore can be significant,
making the application of Darcy’s equation inadequate to
describe the flow. The inertial effect is believed to be the
reason for this non-Darcy flow in porous media (Li and
Engler, 2001) and, due to the higher density, this effect can be
more pronounced for CO2 than natural gas.
Even if the CO2 is at critical or supercritical condition in
the reservoir, as it approaches the wellbore, flows up the
tubing and arrives at the surface it can be, depending of
course, on the prevailing local pressure and temperature, in
any phase (gaseous, solid, liquid or supercritical). A phase
change from the liquid or supercritical state to the gaseous
state (boiling) in the wellbore can results in an unstable slug
flow. This in turn can lead to difficulties in handling the fluid
at surface and in the prediction of the surface pressure and
temperature. At the surface, rapid expansion of CO2-rich gas
downstream of a choke to atmospheric pressure can reduce
temperatures sufficiently, through the Joule-Thompson effect,
that water ice and dry ice are formed.
Over a pressure range of 1-20 MPa, and at a temperature
of 70 oC, CO2-rich gas exhibits a greater range in the gas
deviation factor (z factor) than a typical natural gas (1.0 to
0.36 cf. 1.0 to 0.75). Obviously a z factor correction must be
taken into account in the surface flow rate calculations or else
errors in excess of 60% may occur. Another important
application of the deviation factor, along with the pressure
dependence of the viscosity, is in the calculation of pseudo
pressure. The use of pseudo pressure, rather than the more
convenient but often considerably less accurate p2 approach, is
now standard practice in gas well test analyses.
The Well Test
Test Objectives and Plan Outline
The objectives of the well test were to determine the open
flow potential; quantify the non-Darcy effect; determine the
reservoir permeability and well skin factor; estimate the flow-
based reservoir volume and compare this with the map-based
reservoir volume; and to take samples from the flowing stream
for detailed compositional analysis in the laboratory.
A multiple-rate flow test, along with a drawdown test and
buildup were conducted according to the well test plan
designed by the CO2CRC (Appendix). The multiple-rate test
was conducted at four (4) rates, regulated by altering the
choke size at the wellhead, with each flow period lasting for
two (2) hours. During the flowing periods, multiple fluid
samples were taken on the surface for compositional analysis.
The downhole pressure was measured using two precise
quartz gauges, and the surface pressure was also monitored.
Both downhole and surface temperatures were recorded
throughout the tests. Following the multiple-rate flow test, a
thirty (30) hour drawdown test was conducted.
Operations Summary
On behalf of the CO2CRC Otway Project Limited,
Upstream Petroleum coordinated the completion and testing of
Buttress-1. Buttress-1 well (Waarre C formation) was earlier
identified as a gas well that was predominantly a CO2 source
that could be used for production of CO2 with subsequent
injection into a nearby depleted gas well as part of a CO2
Geosequestration Project.
Buttress-1 well is located in a farming district in the Otway
Basin, Victoria. The well was originally drilled by Santos
Limited in January 2002 to a monobore design with 3.5” 13Cr
95 9.2 ppf Fox production casing. The reservoir of interest is
situated within the Waarre sandstone fairway of the Port
Campbell embayment. Due to the high CO2 gas content in the
Waarre C the well remained unperforated until mid 2006.
In April 2006 a Wood Group Xmas tree, with xylan coated
wetted surfaces, designed for wet CO2 flood service, was
installed. During May 2006 Buttress was drifted, an SCMT
log was run and the casing was pressure tested to 1500 psi
(102 MPa). In June 2006 Buttress was perforated slightly
overbalance over the interval 1615-1619 m RT using a
slickline perforating technique and the Owen PRISM firing
system. After swabbing to induce flow, and then cleaning up,
a modified isochronal flow test was performed followed by a
30 hr extended flow and a 45 hr PBU before suspending the
well.
During the test sampling was conducted by Petrolab. The
reservoir comprises approximately 79% CO2 and 19%
hydrocarbons (Table 1). Sampling also examined process
contaminants. Various studies were undertaken to learn of the
phase behaviour and hydrate forming characteristics of the
fluid for use in later process design.
Prior to commencement of operations the flow testing
arrangement was engineered and a risk assessment and a
HAZOP performed, to address issues including the safety
concern of asphyxiation by heavier than air CO2, noise in a
rural environment, review of hydrate formation potential and
the release of a non-combustible CO2/hydrocarbon mixture
which could not be flared.
Due to the high CO2 content in the gas, the produced gas
was vented into the atmosphere. The flow test arrangement
comprised a 50 m surface flow line terminating with a diffuser
assembly that was buried some 2 m within an acoustic pit
(Fig.1). A flow meter and primary choke manifold (Fig. 2)
was provided immediately downstream of the wellhead. A
secondary choke was installed immediately above the acoustic
pit to control the flow line temperature within safe limits.
Temperatures were monitored and recorded throughout the
extended testing to ensure pipe operating limits were not
exceeded.
SPE 109294
To maintain acceptable noise levels during the gas venting,
a noise attenuation pit was designed into which the gas would
be released from the flow line. The Acoustic pit was designed
to achieve a noise level no greater than 70 dba at 50m. The pit
was filled with scoria, a crushed volcanic rock, which acted as
the noise attenuation medium. At the end of the flow lines
within the acoustic pit was a specially designed diffuser to aid
retention of the scoria. During the well test, the vented gas
dispersed in line with the model, thus proving the value of
dispersion modelling. The acoustic pit was highly successful
with the actual noise levels during the test being even lower
than predicted.
Upstream Petroleum also carried out computer modelling
for CO2 dispersion to ensure that dispersion would result in
safe levels of CO2 at the 50 m safety radius. The CO2 dispersal
was assisted by 15 industrial fans that were designed to
achieve a reduction to less than 3 percent CO2, the short term
exposure limit, within approximately 10m of the flow
discharge point. CO2 monitoring occurred around the
clearance zone and personal monitors were carried by
operations staff.
To prevent soil contamination the Scoria pit was lined to
contain hydrocarbons and other liquids. All wellbore liquids
were collected in the test tank or captured in the lined scoria
pit. A liquids separator was incorporated in the arrangement to
collect the initial fluid to avoid flooding the acoustic pit.
Provision had been made for methanol injection to
suppress hydrates however this proved ineffective and the
separator was bypassed for most of the test. During the clean
up flow, the temperature in the flare line downstream of the
separator dropped to as low as -47°C. To increase line
pressure and maintain normal temperatures backpressure was
held on the flare line using a choke valve located at the entry
to the acoustic pit. Flow rates were estimated using a choke
flow equation.
When the well was shut in, the wellhead pressure built
briefly to 1530 psi (105.5 MPa) before declining to a final
shut-in wellhead pressure of 1446 psi (98.4 MPa). A wellhead
pressure of 1421 psi (96.6 MPa) was measured on August 30.
The pressure decline is thought to be due to temperature
effects as the predominantly CO2 fluid cools in the wellbore
and its density increases with the resultant hydrostatic pressure
increase resulting in a reduced wellhead pressure.
The operations conducted during June 2006 were unique in
that the main objective was to flow a CO2 well, which behaves
very differently to a conventional gas well. The well test
operations were especially challenging, as there have not been
many instances where CO2 wells have been flow tested in the
past. Operations were completed successfully with no
incidents and the required data were obtained successfully.
3
The Test Results, Analyses, and Discussions
Gas Composition and Phase Behaviour
A representative gas composition, based on samples taken
from the flowing stream, is listed in Table 1. With this
composition, the calculated pressure gradient in the reservoir
is 0.148 psi/ft (3.35 kPa/m), which agrees well with the
pressure gradient of 0.146 psi/ft (3.30 kPa/m) based on the
repeated formation test (RFT) shown in Fig. 3.
Pure CO2 has a critical pressure of 7.38 MPa and a critical
temperature of 31.7oC. For the Buttress-1 gas mixture, based
on the Peng-Robinson equation of state, the critical point is
predicted to be 1250 psi (8.5 MPa) and 15.3oC. The laboratory
PVT test results (Fig. 4a, 4b) indicate a critical point of 1250
psi (8.5 MPa), and 10-15oC, which agrees very well with the
equation of state prediction. Notice that the critical point is
different from that of the pure CO2. Impurities in a CO2-rich
gas shift the critical point on a phase plot (Fig. 4c).
During the flow test, the reservoir pressure always
remained above 2,200 psi (14.97 MPa) and the reservoir
temperature held steady at 68 oC. At the wellhead, the pressure
was always above 1350 psi (9.18 MPa), and the temperature
was always above 10oC. As such, the reservoir conditions
were always above the critical point so that the fluid remained
supercritical throughout the well test without undergoing
phase change. At the surface, CO2 were in liquid state during
the shut-in period (about 10 oC, Fig. 5) and supercritical
during the flowing period (about 20 oC). Thus the gas stream
changes phase in the wellbore. Nevertheless, phase change
from supercritical to a liquid state does not involve any
significant or abrupt change in fluid properties. Downstream
of the choke, the gas was vented to the atmosphere, where it
expanded quickly, and formed ice in the disposal pit and also
on the flow line (Fig. 1, 2).
Flow Rate Estimation
The flow rate at the surface, q, is calculated using Eq. 1
(Bradley, 1992), which incorporates the choke coefficient (C),
which is based on the choke size; the pressure (p) and
temperature (T) upstream of the choke; the gas specific
density (γ) and the gas deviation factor (z).
Cp
q= (1)
γTz
Input data and results are listed in Table 2. The calculated
flow rates are different from the designed flow rate
(Appendix) due mainly to the real gas effect i.e. the
contribution of z in Eq. 1, which can cause the flow rate to be
66% larger (z = 0.36) compared to an ideal gas (z =1).
Obviously, the z-factor correction should always be used in
treating a CO2 gas mixture. The fact that design rates were not
met does not in anyway impact on the accuracy and
effectiveness of the well test, and the subsequent analysis.
4 SPE 109294
The Multiple-rate Flow Test Results
Fig. 5 shows the multiple-rate flow test results. The
bottom-hole pressure is plotted over time for the open-and-
shut-in operations. The well quickly reached an apparently
stabilized flowing regime at all four rates (e.g., the stabilized
pressure is about 2,300 psi (15.6 MPa) for the first flowing
period). During the shut-ins, the pressure buildups reached the
initial reservoir pressure very quickly.
During a constant rate drawdown test, the bottom-hole
pressure declines over time (although the decline can be very
slow). In this test, however, the pressure increases slightly
during all four, especially the last three short flow periods
(Fig.5).
It is speculated that the temperature increase at the
wellhead during the flow is responsible for this anomaly. The
reservoir gas temperature is 68 oC, it heats up the wellbore
during flow. Thus, over time the gas loses less heat to the
wellbore, and the gas temperature increases at the wellhead.
Therefore, the actual flow rate through the choke and the
reservoir decreases (Eq. 1), and the bottom-hole pressure
increases. Nevertheless, the pressure increase is not
significant, and an average pressure is taken for each of the
four stabilized flow period in the subsequent calculations.
Calculation of Pseudo Pressure
Using pseudo pressure is now standard practice when
analysing gas well tests. Pseudo pressure is defined as,
−
Δm ( p) = ∫ p 2 p /( μ z ) dp (2)
p wf
where p is the reservoir pressure; pwf is the flowing
bottom-hole pressure; z and μ are the gas deviation factor and
viscosity respectively. This definition is derived from the
fundamental well flow equations. (Note, if the term 2p/(μz) is
linear in p, Δm(p) becomes a function of p2, and if this term
approximately constant, Δm(p) is then a function of p. Pseudo
pressure often takes these convenient forms in certain regions
of the analysis.)
For this gas mixture, at reservoir temperature (68 oC) and
the relevant range of pressure (0 - 20 MPa), z and μ, are
calculated using Eclipse PVTi (Peng-Robinson EoS and the
Lohrenz-Hary-Clark viscosity correlation). The term 2p/(μz) is
then calculated (Table 3) and plotted in Fig. 6. For
comparison this term is also calculated and plotted in Fig. 6
for a hypothetical natural gas (with 80% C1 and 20% CO2).
As can be seen, the curves for the CO2-rich gas and the
natural gas are similar for pressures below about 1,500 psi
(10.2 MPa), but depart from each other as the pressure
increases. (This clearly illustrates the appropriateness of using
the pseudo pressure approach, particularly over the range of
interest of between 2,200 – 2,400 psi, 14.9 – 16.3 MPa). The
curve in Fig. 6 for CO2 can be matched with a parabolic
function (Eq. 3),
2p/(μz) = -0.0309 p2 + 185.75 p – 14572 (3)
so that the pseudo pressure Δm(p) can be calculated using
Eq. 2. The results of this calculation are listed in Table 3.
The Multiple-rate Test Analysis
In Table 4, the stabilized flowing pressure (Fig. 5) for the
four flow rates, Δm(p), Δp2, and Δm(p) /q and Δp2 /q are listed.
The well deliverability curves Δm(p) /q and Δp2 /q versus flow
rate are plotted in Fig. 7.
From the linear deliverability plot shown in Fig. 7(a), the
open flow potential (OFP) of the well can be inferred from the
value when pwf becomes zero, and this is found to be 80
MMscf/d (2.26 MM m3/d). As a comparison, the OFP
prediction from a “p2” approach is shown in Fig. 7(b), which
indicates an OFP value some 20% higher (100 MMscf/d (2.83
MM m3/d )).
Another useful feature of Fig. 7(a) is that it can be used to
estimate the non-Darcy coefficient. For this situation the
coefficient is 1/1.7456 (i.e. 0.575), which indicates a
significant departure from Darcy flow (when the coefficient
would be 1).
The deliverability plot (Fig. 7(a)) can also be used to infer
the permeability thickness product, kh (Beggs, 1984).
Theoretically we have,
Δm(p) = aq + bq2 (4)
where a and b are coefficients, and the second term
represents the non-Darcy contribution to the overall pressure
loss. The coefficients in Eq. 4 are,
a = 1422.4T /(kh) [ln(rd/rw)+S] (5)
b =1422.4T/(kh) [D] (6)
where T is the temperature; kh is the permeability
thickness product; rd is the drainage radius; rw is the wellbore
radius; S is the mechanical skin factor, and D is the non-Darcy
coefficient.
From the data in Table 4, values of a and b are found to be
554.8 and 0.075 respectively. Unfortunately a detailed
evaluation of the permeability and skin factor was not possible
from the data collected but, assuming a skin factor range of
between 5 and 15, the permeability would be between 700 mD
and 1200 mD.
The Drawdown Test
Fig. 8 shows the bottom-hole pressure profile during the
30-hour drawdown flow test. After approximately 15 hours
(Fig. 8(a)), the pressure gradient becomes nearly constant,
indicating the transition from the late transient flow regime to
the semi-steady state regime.
SPE 109294 5

The linear slope during the semi-steady state period (Fig. Nomenclature
8(b) is used to calculate the gas in place that had an influence a = coefficient
on the test. This is performed using Eq. 7, B = formation volume factor
b = coefficient
vp = 0.0418 q B/(βLc) (7) C = choke coefficient
c = total compressibility, 1/MPa
where vp is the contributing pore volume, q equals 5.5 cg = gas compressibility, 1/MPa
MMscf/d (Row 2 in Table 2); B is the formation volume cw = water compressibility, 1/MPa
factor, 0.0048; βL is the slope of the linear plot, 0.063 psi/hr; cf = rock compressibility, 1/MPa
and c is the total compressibility, 0.00057 psi-1. (Appendix) D = non-Darcy coefficient, d/Mscf
h = formation thickness, m
thus, k = permeability, mD
p = pressure, MPa
−
vp = 30.730 MMcf (0.871 MM m3) (8) p = reservoir pressure, MPa
pwf = well bottom hole flow pressure, MPa
and the gas initially in place (GIIP) is then, q = flow rate, m3/d
rd = radius of drainage area, m
GIIP = 6.40 Bscf (181 MM m3) (9) rw = well radius, m
S = mechanical skin factor
Given the general level of uncertainty in the collected data, Sw = water saturation
in identifying the constant gradient and in estimating the total Sg = gas saturation
system compressibility etc. this value for the GIIP can only be T = temperature, C
treated as an indicative numbers. However, it should be noted t = time, hour
that the result shows reasonable agreement with the estimate tD = dimensionless time
made from the mapped reservoir model which has a possible vp = pore volume, m3
gas-water contact at 1706 m, based on pressure survey data, z = deviation factor
leading to an initial gas in place estimate of 5.2 Bscf (147 MM
m3). Greek symbols
μ = viscosity, cP
Due to the high formation permeability and low viscosity
φ = porosity
of the reservoir fluid, the pressure buildup and drawdown tests
βL = coefficient, psi/hr
yielded very few data points in the transient period, making it
impossible to reliably identify this regime of the flow. ρ = density, kg/m3
Moreover, the data are masked by the wellbore storage effect. γ = gas specific density
Thus, analyses of these data were believed not practical and Δm(p) = pseudo pressure, psi2/cP
were not attempted.
SI Metric Conversion Factors
Conclusions cp × 1.0* E-03 = Pa s
Based on this analysis we conclude that: mD × 1.0132 E-15 = m2
1. CO2-rich gas exhibits different flowing behaviours ft × 3.048 E-01 = m
compared to a typical natural gas. inch × 2.54 E-02 = m
psi/ft × 2.262 E+04 = Pa/m
2. The fluid properties need to be properly accounted psi × 6.895 E+03 = Pa
for if accurate data analysis is to be realised.
*conversion factor is exact
3. The sampling and compositional analyses indicate a
CO2 content of 78.9 mol% for the Buttress reservoir
gas. Reference:
Ahmed, T., “Reservoir Engineering Handbook,” Second Edition,
4. The Waarre C reservoir bulk permeability is likely to Gulf Professional Publishing, 2001, ISBN 0-88415-770-9
be in the range of 700-1200 mD.
Beggs, D., “Gas Production Operations,” Oil & Gas Consultants
International INC. and H. Dale Beggs, 1984, ISBN 0-930972-06-6
5. The extended drawdown test indicates a gas initially
in place (GIIP) figure of 6.4 Bscf (181 MM m3), Bradley, H., “Petroleum Engineering Handbook,” Society of
which is in reasonable agreement with an original Petroleum Engineers, 1987, ISBN 1-55563-010-3
map-based estimate.
Jokhio, S., Tiab, D., and Escobar, F., “Quantitative Analysis of
Deliverability, Decline Curve, and Pressure Tests in CO2 Rich
Reservoirs,” SPE 70017, paper presented at the SPE Permian Basin
6 SPE 109294

Oil and Gas Recovery Conference, held in Midland, Texas, 15-16 From tD = 0.000264 kt/(φμcA), we have,
May 2001
t = tDφμcA/ (0.000264 k)
Li, D., and Engler, T., “Literature Review on Correlations of the
Non-Darcy Coefficient,” SPE 70015, paper presented at the SPE
Permian Basin Oil and Gas Recovery Conference, held in Midland, taking,
Texas, 15-16 May 2001
φ = 0.215
Matthews, S., and Russell, G., “Pressure Buildup and Flow Tests in μ = 0.029 cP
Wells,” SPE Monograph Volume 1, 1967, Henry L. Doherty Series. c = cg Sg + cw Sw + cf ≈ 5.7 x 10-4 psi-1
ISBN 0-89520-200-X A = 5.5 x 105 m2 = 5.917 x 106 ft2
k = 2000 mD
Appendix
The Objectives of the Well Test we have,
• Obtaining reservoir properties, well open flow
potential, non-Darcy effect, and well performance t = 18.42 hours
(the well deliverability).
• Obtaining flow-based reservoir volume to compare In this calculation, values of the parameters were chosen with
with the map-based reservoir volume uncertainties. Thus, we recommend a test time of 30 hours for
• Fluid sampling the final drawdown.

Available field/well information prior to the well test Calculation of the kh Product
• Reservoir area ≈ 5.5 x 105 m2 (from the top structure From Eqs. 5 and 6, we have,
map); thickness = 10 m
• Estimated porosity = 0.215, water saturation = 0.25 kh = 1422.4T/a [ln(rd/rw)+S]
(from logs), permeability = 1-3 Darcy (from
correlation) D = b kh/(1422.4T)
• GIIP: 5.2 Bscf (Based on geological interpretation,
i.e., the map, GWC, porosity, water saturation, and where values of a and b are 554.8, 0.075 respectively,
formation volume factor)
T = 614 oR (68 oC),
• Reservoir pressure 2330 psi (15.8 MPa); Temperature
68 oC
rd is the drainage radius,
• Reservoir gas composition from RFT sampling: CO2
= 85%, Methane = 14%
rd = 1.5 (A/CA)0.5

Perforation interval
• Perforation targets Waarre C formation
• perforation interval: 1615-1620 m (Logger’s depth,
MD), see Fig. A1.
The rational is to perforate the high-quality sand at the
upper of the formation to reduce the risk of water production.
Multiple-rate Test
A multiple-rate test is recommended to meet the test
objectives. Detailed steps are illustrated in Fig. A2. The test
includes 5 flowing periods and 4 build-up periods. The
recommended flow rates are 2, 5, 8, 10 MMscf/d, and the
drawdown period is 3 MMscf/d. Each of the first 4 flowing
period lasts 2 hours, and the last flowing period lasts at least
30 hours to ensure reaching of the pseudo-steady state
condition.
where A is drainage area = 5.5 x 105 m2 = 5.917 x 106 ft2; CA
is the shape factor to account for the effect of well/field
configuration (Fig. A3). A value of 2.077 is assigned to CA
(Matthews and Russell, 1967); rw is the wellbore radius, which
is 3 ½’’ (0.292 ft); S is the mechanical skin factor.
(Unfortunately without high-quality buildup data, the skin
factor cannot be determined); D is the non-Darcy coefficient.
For skin factor to be 5-15 (including possible geometry skin
since the formation is only partially perforated),
kh = 23462.08 - 39190.1 mD ft
The thickness of the reservoir sand, h, is about 10 m, thus, the
permeability would be 715 mD to 1194 mD, and the non-
Darcy coefficient is,

In the extended drawdown period, to ensure the test data D = 0.002 - 0.00336 d/Mscf
reaches semi-steady state condition, it requires that the
dimensionless time for the drainage shape (Fig. A3), Calculation of the Reservoir Volume
The contributing pore volume, vp, is given by
tD > 0.5
vp = 0.0418 q B/(βLc),
SPE 109294 7

where Table 2. Flow rate calculation

Choke Choke Pre. psi Temp. Gas z Flow rate
q, flow rate, is, 5477.83 Mscf/d (Table 2, Row 2); size Coeff. o
gravity Mscf/d
R

B, the formation volume factor, is given by (Ahmed, 2001) 5/16 ” 43.64 1502 530 1.33 0.36 4114

3/8 ” 61.21 1448 532 1.33 0.37 5477

B = 0.02827 pT/z
= 0.002827 x 2300 x 614/0.622 7/16 ” 85.13 1485 533 1.33 0.38 7702
= 0.0047 9/16 ” 144.1 1422 535 1.33 0.41 11996

11/16 ” 218.96 1374 537 1.33 0.44 16971

where
p is pressure in psi,
T is temperature in oR Table 3. Gas mixture properties and the pseudo-
z is gas deviation factor. pressure function
μx
c, the total system compressibility, is given by, 0.0001 2p/(μz),
Δm(p), psi /cP
2
p, psi z cP psi/cP
c= cg Sg + cw Sw + cf 2940 0.616 371 257039.52
= 0.00053 x 0.75 + 0.25 x 3 x 10-6 +4 x 10-6 2793 0.614 354 256997.74
= 0.00057 psi-1 2646 0.614 336 256514.66
2499 0.616 319 254345.97
1 1 ∂z
where cg = − [ ] (and can be calculated based on
2352 0.621 301 251657.12
p z ∂p 2316 0.622 297 250738.90 0.00

data in Table 3) 2306 0.622 295 250985.68 2493210.64

2296 0.623 294 250546.16 4982112.35
βL, the linear pressure decline rate during the semi-steady state 2286 0.623 293 250105.72 7466643.35
period, is 0.063 psi/hr 2276 0.624 292 249624.36 9946741.82
2266 0.625 291 249182.13 12422345.98
Thus, 2256 0.625 289 249559.73 14893394.02
2246 0.626 288 249117.11 17359824.13
vp = 30.730 MMft3 (reservoir volume) 2236 0.626 287 248633.82 19821574.53
2226 0.627 286 248149.75 22278583.40
so that the gas initially in place (GIIP) is then
2216 0.627 285 247664.87 24730788.96

GIIP = vp/B = 6.40 Bscf 2205 0.629 283 247743.07 27422593.48

2058 0.639 267 241247.74 62789474.34
Table 1. Composition of a Buttress-1 gas samples 1911 0.654 251 232830.15 96894052.51
Component Mol% 1764 0.672 236 222457.63 129540018.85
Hydrogen Sulphide H2S 0.00
1617 0.693 224 208333.33 160531064.26
Carbon Dioxide CO2 78.72
Nitrogen N2 1.41 1470 0.718 212 193146.58 189670879.59
Methane C1 18.76 1323 0.741 203 175903.95 216763155.74
Ethane C2 0.51
1176 0.768 195 157051.28 241611583.58
Propane C3 0.24
Iso-Butane iC4 0.03 1029 0.795 189 136967.16 264019854.00
N-Butane nC4 0.04 882 0.824 183 116982.33 283791657.86
Iso-Pentane iC5 0.04 735 0.853 179 96275.39 300730686.05
N-Pentane nC5 0.02
Hexanes C6 0.03 588 0.882 175 76190.48 314640629.45
Heptanes C7 0.07 441 0.912 172 56227.05 325325178.94
Octanes C8 0.04 294 0.941 170 36756.89 332588025.39
Nonanes C9 0.04
Decanes C10 0.04 147 0.971 168 18022.66 336232859.68
Undecanes C11 0.01 14.7 0.997 166 1776.41 336257560.66
Dodecanes Plus C12+ 0.00
8 SPE 109294

Table 4. Data for deliverability analysis

2 2 Pressure, psi
Press Flow p - Δm(p)/q (p -
Δm(p),
2 2 2
ure, pwf rate, q pwf , , psi pwf )/q, 2330 2331 2332 2333 2334 2335
2 2 2
(psi) Mscf/d psi /cP psi Mscf/d/cP psi Mscf/d 5120
2316 ( p )

Depth, ft SS
2300 4110 2353100 73856 572 0.0179 5130
2269 7700 10811375 215495 1404 0.0279
2233.5 12000 20367363 375333 1697 0.0312 5140
2208 16970 27145800 488592 1599 0.0287
5150

5160

Figure 3. RFT pressure measurements at Buttress-1 indicating a

pressure gradient of 0.146 psi/ft in the gas column, agrees with a
calculated value of 0.148 psi/ft based on the composition

Figure 1. Test site: CO2-rich gas is flared into the acoustic pit. The
industrial fans were used to disperse the CO2

Figure 4a. PVT test results: pressure versus temperature

Figure 2. The main choke, ice formation outside of the

downstream of the choke

Figure 4b. PVT test results: liquid volume versus pressure at

o
temperatures of 5, 10, 15 and 20 C
SPE 109294 9

1000000000

y = 1.3733x1.7456

Δ m(p), psi /cP

100000000

2
10000000

1000000
1000 10000 100000 1000000

Flow rate, M scfd

Figure 4c. The locus of the critical point for the CO2/CH4 mixture Figure 7. (a). Deliverability curve using Δm(p)

10000000

50
2320

2
45

p - p wf , psi
2280 40
1000000
Temperature, C

2
2240 35

2200 30
P wf , psi

25
2160 100000
20
2120
15 y = 1.0509x1.3523
2080 10
2040 5 10000
1000 10000 100000 1000000
2000 0
0 5 10 15 20 Flow rate, M scfd
T ime, hours Figure 7(b). Deliverability curve using Δp
2

Figure 5. The multiple-rate flow test results

2295

350000 2294
Pressure, Psi

300000
CH 4
250000
2293
2p/( μ z)

200000
CO 2 2292
150000

100000
2291
50000 0 5 10 15 20 25 30
0
0 500 1000 1500 2000 2500 3000 Time, hrs
Pressure, psi Figure 8(a). pressure drawdown during the extended flow period

Figure 6. Comparison of pseudo pressure calculation for the

Buttress CO2-rich gas and a natural gas
10 SPE 109294

q Long flow 30 hrs

2293.0 short flows 2 hrs

2292.5
Pressure, Psi

2 2 2 2
2292.0

2291.5
pw

2291.0
15 20 25 30

Time, hrs

Figure 8(b). pressure drawdown during the semi-steady state Time, hrs
regime.
Figure A2. The Modified isochronal test steps

Waarre C

Perforation
Interval

Figure A3. Buttress-1 well and drainage area configuration

Figure A1. Buttress-1 well log chart, perforation interval