Professional Documents
Culture Documents
SPE 101099 Burning Reserves For Greater Recovery? Air Injection Potential in Australian Light Oil Reservoirs
SPE 101099 Burning Reserves For Greater Recovery? Air Injection Potential in Australian Light Oil Reservoirs
SPE 101099 Burning Reserves For Greater Recovery? Air Injection Potential in Australian Light Oil Reservoirs
Burning Reserves For Greater Recovery? Air Injection Potential In Australian Light Oil
Reservoirs
B.L. Hughes* and H.K. Sarma, SPE, U. of Adelaide
* Now with Sarawak Shell Berhad
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
Introduction
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to Air injection was first introduced as a secondary recovery
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 technique in 19791 (Buffalo Field) to improve upon primary
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
recovery in deep, high temperature, low relief, low
for commercial purposes without the written consent of the Society of Petroleum Engineers is permeability reservoirs. Since its first application, air injection
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 has been applied successfully, technically and economically as
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
both a secondary and tertiary (Double Displacement Process,
DDP, West Hackberry2,3) EOR process, over a variety of
Abstract reservoir scenarios, in both vertical and horizontal flooding
Air injection is an Enhanced Oil Recovery (EOR) technique modes. Full-field air injection projects have been developed as
with limited exposure in the Asia-Pacific region and no recently as 19964,5 (Horse Creek), and a range of field pilots
previous application in Australia. Analogy with successful air and evaluation studies are underway or have been proposed in
injection projects in the USA, suggests that it could be a recent times; these have tended to focus on the use of the
suitable EOR process for onshore light oil fields in Australia; process for tertiary recovery, including the potential of the
no evaluation has been conducted to date. process for use offshore in the North Sea6,7.
Using open file data, high level screening criteria are used One example of the successful application of air injection
in this study to identify prospective petroleum basins, and an for secondary EOR is the Medicine Pole Hills (MPH) field1,8.
individual candidate reservoir is examined through a This field is indicative of those projects which have been
simulation study. Key issues in the application of the undertaken in the Williston Basin in the Dakotas. The
technique are discussed, as are directions for implementation Williston Basin has been the focus of air injection for
in Australia. secondary recovery for over three decades. The reservoirs in
Air injection involves the continuous injection of high- the Williston Basin (Buffalo, MPH, Horse Creek) are low
pressure air into the reservoir. The oxygen in the air reacts permeability non-fractured carbonate reservoirs. There is little
with the reservoir crude, consuming 5-10 % of the Original- structural relief, and the primary drive mechanism of liquid
Oil-In-Place (OOIP) and generating flue gases in-situ. This and rock expansion results in low primary oil recoveries due to
creates a gas drive process and acts to re-pressurize the a lack of pressure support. This results in rapidly declining
reservoir. The process does not require water as a mobility production rates, and a high residual oil saturation due to
control agent; a significant advantage in water-scarce capillarity. The air injection technique allowed for rapid
Australia. It could also replace hydrocarbon (HC) miscible reservoir re-pressurisation, to provide energy, and helped
floods, freeing cleaner HC gases for energy use. Ideally the mobilise trapped oil, through improved sweep efficiency.
process is suited to deep, high-temperature, light oil reservoirs, Air inejection was initiated at MPH in 1986, but only
and is applicable to both secondary and tertiary recovery. continued for two months before injection was stopped due to
The Cooper-Eromanga Basin, Carnarvon Basin (Barrow a decline in oil price. Injection recommenced in October 1987,
Island) and the Surat-Bowen Basin were identified as the most and continued uninterrupted. As of June 1994 the oil rate at
prospective. The simulation study conducted for ‘Reservoir A’ MPH was 950 bbl/d as compared to 400 bbl/d prior to air
in the Cooper Basin indicated the potential for spontaneous injection. The number of producing wells has remained
ignition and propagation of a stable combustion front within constant throughout field life, but early in field life a series of
the reservoir; hence it is a potentially good candidate for EOR acid fracturing jobs was performed. These jobs have therefore
by air injection. made a small contribution to the production rate in early life.
Given the ‘high’ oil price and maturity of Australia’s oil The GOR (Gas-Oil-Ratio) has consistently increased as
provinces, significant value is associated with EOR. Air injection has continued; this suggests rapid breakthrough of
injection is potentially suitable for Australian onshore flue-gas. The increase in GOR has been matched by an
2 SPE 101099
increase in the Natural Gas Liquid (NGL) content. An NGL from a process point of view that a large and reliable injection
recovery plant was installed and during the period 1991-1993, gas source is required. In order to support the injection of large
NGL recovery was approximately 200 bbl/d. The produced gas volumes, which would be required for an offshore project,
gas stream at MPH consists of CO2, N2, hydrocarbon and trace potentially significant infrastructure changes (‘topside’
gases. This indicates that oxygen is being consumed by modification) i.e., addition or modification of facilities e.g.
reaction within the reservoir. compression, purification, would be required. On an existing
Other fields and their performance are not discussed here facility for which these changes were not included during
as the focus of this paper is not to present a review of design, such changes could be structurally or cost prohibitive.
existing/previous projects. Nevertheless, reservoir properties Air provides a significant advantage in this environment
for previous and possible/current air injection projects and because it is available without transport infrastructure, and
project performance data for previous/current projects are only requires compression facilities. It may therefore have
presented in Tables A-1 and A-2 respectively. more potential than other processes on existing facilities with
Air injection was one of a number of potential EOR limited margin for modification. The major hurdle to be
techniques considered in all projects in which it has been used. overcome with the offshore use of air injection is the safety
Alternate techniques included water injection, gas injection, hazards associated with having high-pressure air and
and horizontal wells. Key reasons for which air injection was hydrocarbon in close proximity. If this issue can be addressed
chosen over alternate techniques are, satisfactorily, then air injection has the potential to have a
• superior injectivity of air in comparison to water, major impact on secondary and tertiary recovery offshore.
• faster re-pressurisation than with water; air The potential benefits of air injection when applied to
compressibility, fields in Australia, in combination with strong analogy to
• preference for gravity drainage rather than pushing reservoir characteristics in the successful Williston Basin
oil up structure due to more favourable relative projects provide a strong motivation for further evaluation of
permeability effects (West Hackberry2,3 ), air injection as an EOR technique for Australian light oil
• cost benefit compared to other injectants; CO2, N2, reservoirs
hydrocarbon gas,
• previous technical and economic success in nearby Air Injection Theory
fields (Williston Basin). The air injection process involves the continuous injection of
high-pressure air (79% N2, 21% O2) into the reservoir. The
The use of the process at West Hackberry2,3 (salt-dome, high oxygen reacts with the reservoir crude, preferably by
relief, high permeability, high secondary recovery from water combustion consuming approximately 5-10% OOIP, and is
flood), represents the proof of the air injection process for itself consumed. The combustion reactions release heat and
tertiary recovery in high permeability reservoirs. In the generate carbon oxides (CO2, CO) producing ‘flue-gas’ (CO2,
Williston Basin, air injection was selected because it best met CO, N2. In light oil reservoirs, initiation of reactions in the
the technical and economic requirements. In the West reservoir is typically spontaneous due to the small activation
Hackberry project, the selection of air over other gas injectants energy. An artificial ignition source is therefore not required.
was based upon purely economic criteria, taking advantage of The aim of air injection in light oil reservoirs is not to
technically suitable reservoir conditions to use air as a more generate heat and promote EOR by thermal effects; rather, to
cost effective injectant. The general advantages/disadvantages create a gas drive by generation of flue-gas in situ. It is
of air injection are summarised below in Table 1. envisaged that gas drive will be the dominant mechanism, at
least in the earlier phase or when the ignition front is far
Table 1: Air Injection: Advantages & Disadvantages removed from the production well.
Advantages Disadvantages The process operates effectively as both a miscible and
• Does not consume • Compressors must be immiscible gas flood. This means that whilst the reaction
saleable gas for EOR specially fabricated pathway (combustion) remains critical and that heat generated
• Does not require source • Doesn’t attract does contribute to recovery (if the combustion front is pushed
of CO2 sequestration benefits of
• Does not require water CO2 from the injector all the way to the producer) the temperature
either for EOR or as a • Has not been proven range in which the reaction occurs, and, hence the degree of
stability buffer (WAG) as offshore exothermicity is not critical to an efficient recovery process.
for other gas processes Effective recovery can occur provided all oxygen is consumed
• Superior injectivity to
water
regardless of the temperature regime, Low Temperature
• Can be used in super- Oxidation (LTO), 150-350°C or High Temperature Oxidation
wet fields (HTO) 450°C+. It is imperative that all oxygen is consumed;
• Unlimited, universally for both safe operation of the process and with regard to the
available ‘free’ injectant
potential for the microbial degradation of the oil if free oxygen
was present in the reservoir.
Although air injection remains an un-proven technique The reaction of oxygen with a reservoir crude is an
offshore, it has received a great deal of interest for offshore extremely complex one. The reaction pathway of a crude is
application, in recent times. One logical driver for this is that determined by compostion, reservoir temperature, pressure,
generally offshore developments are significantly larger than rock mineralogy and rock heat loss characteristics. The heat
onshore. This in effect means that the ‘prize’ is bigger, but
SPE 101099 3
loss characteristics are in turn affected by the porosity. In a A reservoir undergoing air injection can be broken down into a
light oil the desired reaction path of combustion is sustainable series of zones according to the processes and temperature of
(provided the oil has sufficient fuel load) in either the HTO or the zone. The schematic shown in Figure 1 describes these
LTO regime. In light oil, LTO is often the favoured reaction zones when the process is operating under ‘bond-scission’
pathway9 and combustion under HTO may not be sustainable mode. The specific mechanisms active in the air injection
given the fuel load of the oil. process will depend on exactly which mode i.e. miscible vs.
Combustion is a ‘nice’ and imprecise way of describing immiscible and the temperature range the process is operating
the reaction pathway; combustion represents the overall under.
reaction. The complex reaction ‘system’ is generally
represented in two parts, oxidation (addition) followed by
‘decomposition’6/’bond-scission’9. The first reaction describes
the formation of oxygenated compounds and the second the
breakdown of oxygenated compounds into carbon oxides and
a long chain hyrdocarbon. The other reaction, which occurs in
air injection, is cracking. In this case large molecules are
broken down into one lighter and one heavier component due
to heat. Operation in bond-scission mode i.e. combustion
rather than only addition is critical. Under addition,
oxygenated compounds will tend to react by polymerizing
with each other; this increases the oil viscosity. Removal of
the O2 from the gas stream without replacement by flue-gas to
maintain pore pressure promotes trapping of the oil phase9.
Based on studies of light oils, addition reactions tend to
dominate below 150°C9 (temperature under reaction not static
reservoir temperature).
Injector Producer
1. Swept Zone
O2 Concentration
20 2. Combustion Front
3. Oil & Steam Banks
10 4. Unswept Reservoir
1 2 3 4
Temperature Profile
Reservoir Temperature
NB* The combustion front is very thin compared to the other zones,
but is horizontally exaggerated in this schematic.
Process Recovery Mechanisms potential areas was limited to onshore because air injection
The recovery mechanisms which may be attributed to air remains unproven offshore.
injection are summarized below,
• improved sweep efficiency due to flue-gas sweeping,
• rapid re-pressurization of the reservoir,
• ‘stripping’ of light ends (volatiles) by flue-gas and
subsequent NGL recovery, Carnarvon
Basin,
• oil swelling by flue-gas, Barrow
• injection gas substitution due to the generation of Island
Amadeus
CO2 in-situ, Basin
• miscibility of flue-gas with oil, Surat-
• operation above the critical point of water and Bowen
Basin
therefore potential super-extraction benefits
associated with steam fingering ahead of the
Cooper-
combustion front resulting in viscosity reduction and
Eromanga
oil mobilization. Basin
potential candidates are identified in the basin-by-basin reservoirs17 which have similar temperature, fluid saturation,
discussion. and oil API characteristics has demonstrated these reservoirs
to be potential candidates for the process.
Amadeus Basin The question of miscibility affects the efficiency of the
The Mereenie reservoir in the Amadeus Basin was considered flue-gas drive. A miscible process will be more efficient and
as a potential air injection target. General reservoir properties recover more oil. The reservoirs of Barrow Island are quite
appeared to be appropriate but there is a large gas cap, which shallow, and have fairly low initial pressures. Therefore, if
is being actively produced as part of the production strategy. Minimum Miscibility Pressure (MMP) could not be obtained,
Therefore, whilst the reservoir parameters may be appropriate, the process would operate in the less efficient immiscible
the use of air injection for EOR is not compatible with the flood mode. This would still allow for better sweep, although
field production strategy. No further evaluation of the generating less significant volumes of incremental oil.
Mereenie field was carried out.
Cooper-Eromanga Basin11, 12, 13, 14, 15
10
Barrow Island Reservoirs in the Cooper-Eromanga Basin were identified as
Barrow Island is the only field to be evaluated for the potential potential targets for air injection because of similarities
of light oil air injection in the Carnarvon basin. This is the between some reservoirs in this region and the successful
only field to be evaluated because of the logistical access that projects of the Williston Basin i.e. high temperature, low
Barrow Island provides i.e. the reservoir is effectively permeability reservoirs with low primary recovery. The
onshore. Eromanga Basin overlies the Cooper Basin forming a
The Barrow Island structure is a broad anticline with ‘stacked’ basin. Oil is produced from both units and often
multiple oil and gas producing horizons, and an oil-in-place of from the same field. There is potential in both units for the
1250 MMbbl. The most significant oil producing zone is the application of air injection.
Windalia Sand, which has produced the bulk (95%) of the 278 The oil-bearing reservoir units of the Eromanga Basin are
MMbbl produced to 1999. Owing to poor reservoir quality in the Murta Formation, McKinlay Member, Namur Sandstone,
the Windalia Sand, water flooding has been used in the field Birkhead Formation, Hutton Sandstone and Poolowanna
since 1968 to enhance oil recovery. Facies. The basic reservoir properties of these units are
The other reservoirs which contribute to the bulk of provided in Table B-2. Fields which broadly represent the oil
production not from the Windalia Sand are the M3 Zone, the producing fields in the Eromanga Basin are,
Mardie Greensand, the Tunney member of the Muderong • Strezlecki - producing from McKinlay Member-
Shale, and the Flacourt and Dupuy formations. Other oil zones Namur Sandstone, intra-Namur Sandstone, Birkhead
which produce only minor amounts are the Gearle Siltstone, Formation-Hutton Sandstone,
M4 Zone and Malouet Formation. A full set of available • Dullingari – producing from Murta Formation,
reservoir data for these oil bearing zones is included in Table McKinlay Member-Namur Sandstone,
B-1. • Jena - producing from the Murta Member, McKinlay
The aim of air injection into the Barrow Island reservoirs Member,
would be to improve sweep efficiency and mobilize the • Keleary – Poolowanna Facies.
bypassed oil. The variable nature of the permeability in
Barrow Island to both extremes suggests that production by The McKinlay Member, Namur Sandstone and Hutton
either water flood or water drive could be improved. Air Sandstone reservoirs tend to contain large oil pools.
injection would give better injectivity and allow the recovery The main oil producing reservoir in the Cooper Basin is
of bypassed oil. Based on the reservoir properties of Barrow the Tirrawarra Sandstone which comprises Permian braided
Island, potential issues with the air injection process are the delta and braided stream sediments. The bulk of oil (80%)
thickness of the reservoir, reservoir temperature, water produced from the Tirrawarra Sandstone is produced at the
saturation, and degree of miscibility given the shallow depth. Tirrawarra field, with Fly Lake, and Moorari-Woolkina
The thickness of reservoir units in Barrow Island ranges contributing the remaining oil. Significant oil accumulations
from 50-130 ft. Given that the structure has low relief, there is have also been identified in the fluvial sandstones of the
significant potential for gas override which would result in an Triassic Nappameri group which are produced by the Telopea
inefficient and potentially unsafe process. There is some and Keleary fields (Eromanga Basin).
heterogeneity within the reservoir; tight sandstone and The Tirrawarra and Moorari fields exhibited low primary
siltstone layers which act as vertical flow barriers within the recovery, and the need to supplement the depletion drive
Windalia Sand may provide ‘thin’ enough flow layers to make recovery was recognised early in field life. The reservoirs
the process stable. Some of the other less prolific thinner sands were studied and miscible gas injection was determined as a
may also be appropriate under this condition. suitable EOR technique. A miscible gas flood pilot was started
The reservoir temperature in Barrow Island ranges from in 1984 with separator gas from the Patchawarra Formation,
46˚C to approximately 100˚C, but in several reservoirs is and ethane injection began in Tirrawarra in 1985. In 1996,
below 75˚C. Considering this in addition to the high water ethane injection into both fields ceased; due to increased gas
saturation which causes substantial heat loss, the question is and ethane demand. No other gas injection process has been
whether the process would auto-ignite and be self-sustaining. utilised since this date. CO2 injection has been considered for a
This question can only be answered by experimental testing. miscible flooding process, but was considered to be too
However experimental testing on the Avilé and Tronsco
6 SPE 101099
expensive. The Reservoir A field which provided the base data heterogeneity, and a low recovery factor and therefore under-
set for the hypothetical simulation study is also located in the developed.
Cooper Basin, and the study presented by Bon and Sarma18 Much of the data from the Moonie field is very old and
suggests it is also amenable to a miscible CO2 flood. Table B-3 presents a few basic data. The reservoir zone in the
The reservoirs of the Cooper-Eromanga Basin are Moonie field is the Precipice Sandstone. The reservoir is
generally good candidates for air injection. The reservoirs are divided into two main sand units, the 56-4 sand which is
hot and quite deep with sufficient residual oil saturation. In below the Evergreen Shale which acts as a seal, and the 58-0
addition, whilst other EOR processes such as CO2 flooding Sand (Lower Precipice Sandstone) which is located below the
may be suitable, air injection may provide a more cost 56-4 Sand. The 56-4 and 58-0 sands are separated by a thick
effective alternative, provided the fuel load in the oil is (30 m) shale formation. Within the 58-0 Sand there is
adequate to sustain ignition. The applicability of miscible CO2 generally good horizontal permeability, but effectively zero
flooding strongly suggests that air injection would be an vertical permeability due to tight streaks and shale barriers.
appropriate EOR technique, and should certainly be The 56-4 Sand is a tighter sand and is not productive in all
investigated further, particularly with regard to continuing the wells due to varying contacts. Reports from written sources
EOR process at Tirrawarra and Moorari. suggest that recovery from the 58-0 sand is very high (93%),
The only concern for implementation of air injection in the but the picture that emerges from discussion with industry
Cooper-Eromanga is the sand thickness given the low relief of professionals suggests that recovery has been far less efficient.
the structural elements of the basin, which could result in gas The authors believe that it provides limited consideration of all
override. The reservoir units of the Cooper-Eromanga Basin potential units, i.e. the less desirable 56-4 sand is not fully
generally show significant permeability variation and tend to incorporated in the volumes. From the perspective of applying
have quite low average permeabilities. The degree of air injection, the 56-4 Sand would, on the basis of available
heterogeneity is likely to be sufficient to ensure that the information be the target for the process.
process would be stable. The Precipice Sandstone has significant heterogeneity in
the form of shale layers and tight streaks; hence large-scale
Paleo-oil Recovery gas override is unlikely. The reservoir temperature and
‘Paleo-oil’ is the residual oil which remains trapped in the relatively high water saturation raise the question of whether
transition zone (below the Oil-Water-Contact) after the the crude oil has the potential to auto-ignite, and whether a
upward displacement of the oil column. In large fields the stable combustion front could be maintained. This is similar to
volumes of paleo-oil can be significant despite the low the consideration of reservoir temperature and water saturation
residual oil saturation and high water saturation of the in Barrow Island. Again, experimental testing is the only way
transition zone. Gas based EOR techniques either in miscible to determine whether the process would be applicable.
or immiscible mode have potential in recovering paleo-oil.
Significant volumes of paleo-oil have beem identified in Basin Analysis Summary
Santos’ Gidgealpa field within the Hutton sandstone. A CO2 Based on the foregoing discussions, it is evident that there is
gas flood has been identified as a potential recovery technique, clear potential for air injection in Australian light oil
and would likely operate as a near miscible flood. The fact that reservoirs, most notably in the Cooper-Eromanga Basin. The
the flood is only likely to operate in a near-miscible mode Surat-Bowen Basin and Barrow Island (Carnarvon Basin)
means that the benefits of pure CO2 may be reduced. In this show somewhat less potential. Experimental testing is required
case air injection could be a competitive option. to further evaluate the potential for air injection in these areas.
The key question in relation to the use of air injection for As a priority, it is recommended that air injection be evaluated
paleo-oil recovery is whether given the high water saturation, as a replacement EOR process for ethane injection in the
combustion would occur and proceed stably under reservoir Tirrawarra & Moorari fields.
conditions. The operation of the process in this case would be
similar in terms of saturation to ‘super-wet combustion’ as Hypothetical Simulation Study
described in In-Situ Combustion (ISC) literature. Experimental A hypothetical simulation study was carried out using the
testing would be critical to determining whether air injection CMG STARS thermal, 3-phase, multi-component simulator.
could be a potential EOR method for this type of recovery. The aim of this simulation study was to further evaluate the
The authors are unaware of any previous application of air potential of a light oil air injection candidate using an
injection to recover paleo-oil, but believe that there is hypothetical data set.
significant potential due to the economy and availability of air The data set chosen from this study is that of ‘Reservoir
as an injectant. This would have a particularly large impact A’18 a light oil reservoir in the Cooper Basin which is
offshore where CO2 has already been considered for paleo-oil prospective for secondary EOR by CO2 flood. This data set
recovery in the North Sea19. was chosen because the reservoir has good prospectivity for
air injection and had the largest and most appropriate data set
Surat-Bowen Basin16 based on open file data. The process used to perform the
The Moonie field, Australia’s first commercial oilfield20, in simulation is described by Figure 3.
the Surat Basin was identified on a similar basis to that of the
Cooper-Eromanga Basin; an onshore reservoir with significant
SPE 101099 7
Key Limitations those for Reservoir A. Simulating any reservoir without a full
There are two key limitations of the model used in the set of accurate reservoir data, or production history places a
simulation of the air injection process for Reservoir A, significant limitation on the model. Clearly, making
• assumed reaction kinetic data (experimentally assumptions for unavailable data can decrease the accuracy of
measured for a previous air injection project), the simulation, even if reasonable assumptions are made. A
• limited open file data available without a full set of lack of production data for history matching means that there
reservoir data or production data for history is no ‘control’ on whether the behaviour of the model is
matching. correct. In this case, data for history matching was unavailable
but was also beyond the scope of the simulation due to time
Reaction kinetic parameters are unique to a given reservoir, restrictions.
crude oil, and in-situ conditions. The factors which affect the Figure 4 shows the data which was available from open
way a reaction occurs are numerous and their contribution is file and that which was assumed to create the hypothetical data
complex. For this reason there is no simple correlation which set used in the model. A value of relative importance has been
can be used, even as a guide to generate reaction kinetic assigned, according to the potential impact of the data, on the
parameters for a set of reactions. The only way of obtaining ‘basic’ i.e. key indicators of the simulation.
these parameters is by experimental testing.
Reaction kinetic parameters, which have been
experimentally obtained for a different air injection project are
used in this study in the absence of any site-specific reaction
kinetics data. This must be recognized as a significant
assumption. For this reason, it would not be unreasonable to
comment that the simulation is flawed, and that it cannot
represent with any degree of accuracy the air injection process
in Reservoir A. However, in the absence of experimentally
determined values for Reservoir A, the previously measured
experimental values are more ‘realistic’ than simply assuming
values. These values provide at least a representative starting
point, and values which are likely to bear some similarity to
b
b
b
10.00%
LOW
5.00%
Figure 4: Reservoir A; Relative Importance of Available &
Assumed Data 0.00%
16
58
100
142
184
226
268
310
352
394
Simulation Grid
The simulation grid used in the model was a 320 acre, quarter Molecular Weight (g/m ol)
of a five-spot pattern, with grid dimensions 3740x3740 ft, cell
Stream MW: 70.53 g/mol
dimensions of 220x220 ft, and a single layer of thickness 50 ft.
An areal Local Grid Refinement (LGR) of 9:1 was applied to Stream MW (Generated Fractions): 68.84 g/mol
those cells adjacent to the injector to facilitate monitoring of
the combustion front.
Figure 5: Generation of Unknown Components
Simulation Reaction Model
The reaction model used in the simulation was designed to The stability of the process in the model does not appear to
match the available reaction kinetic parameters and Reservoir have been affected by the use of volatile component only;
A composition. The composition available from the Reservoir although this could only be rigorously verified by comparison.
A data set is limited to non-hydrocarbon gases, individual
components for methane to hexane, and heptanes plus. Key Variables for Simulation
The reaction model used is based on the combustion Given the limitations on data, and the required assumptions to
reaction of two pseudo components C6-C7, and C8-C17 for create a full data set for simulation, it would have been
which reaction kinetics were available. The model does not unreasonable to expect the simulation to predict perfectly the
include a two-step reaction model i.e. oxidation followed by performance of the process. The aim of the simulation was to
decomposition, and considers the reaction of these two focus on a set of key variables, and to determine whether these
components in the gas phase only. provided a favourable indication of the process. The most
The reaction model was identified as a key limitation of important question to answer initially is whether the oil would
the simulation model, when reviewed by an experienced combust within the reservoir, and if so, whether the
STARS simulator user. The reason for this is that the reaction combustion front formed would be stable. It is also of primary
model does not include any components reacting in the liquid importance to consider whether all the injected oxygen is
phase. The largest component defined in the model is the C8- being consumed and how the oxygen is moving through the
C17 pseudo component; this is a volatile oil. All oils contain reservoir. The variables selected as primary indicators were,
heavier components regardless of their specific gravity, and • Temperature,
these are what ‘drive’ the combustion process. A volatile • Pressure,
component could evaporate, and hence ‘blow-out’ the • gas concentrations, CO2, O2, and N2.
combustion process, resulting in an unstable and inefficient
combustion front, and is therefore insufficient to model the Results
process. A more appropriate set of pseudo components was Results for Temperature and Pressure are displayed in Figures
proposed21 in order to include a liquid reaction component. An 6 and 7. Screenshots of the simulation grid are shown in
extra pseudo component, C18-C30+ was created. Figures D-1 to D-6 for N2, O2, and CO2, concentrations, at the
In order to generate a new pseudo component a start and end of air injection in the simulation.
distribution of the hydrocarbon mole fractions was created. In
this way an individual fraction for each of the C7+
components, (in this case C7- C30+) is determined. The
stream molecular weight of the hydrocarbon is calculated
SPE 101099 9
Discussion
Tem perature and Pressure at Injection Well From the temperature results displayed in Figures 6 and 7,
350 5000 two key observations can be made. Firstly, that a short time
after air injection begins there is an increase in temperature at
the injector (Fig. 6), this represents spontaneous ignition of the
325 4000
Temperature (°F)
Pressure (psia)
temperature zone moves from the injector towards the
300 3000 producer (Fig. 7), indicating that a combustion front has been
initiated and that it is propagating stably through the reservoir.
275 2000
The pressure results confirm that spontaneous ignition has
occurred in the reservoir. Shortly after air injection begins the
pressure at the injection well exhibits a ‘spike’ as shown in
250 1000 Figure 6. A pressure spike is generally associated with the
Dec-03
Dec-07
Dec-11
Dec-15
Dec-19
Nov-23
combustion of reservoir fluid and generation of CO2 in-situ.
The other observation which can be made is about the
Tim e (m onth-year) Temperature
safety and desired operation of the process. From the
simulation screen-shots shown in Figures D-1 to D-6, it can
Pressure (psia)
be seen that whilst flue-gas breakthrough to the injector occurs
Figure 6: Simulation Results: Temperature and Pressure at (Fig. D-2), there is a zero concentration of oxygen ahead of
Injection Well vs. Time. the combustion zone inferred from the region of elevated
temperature (Fig. D-4). Furthermore the oxygen concentration
is restricted to the area behind the elevated temperature front,
Linear Front Distance from Injector
again indicating the location of the combustion front.
800
Dec-07
Dec-11
Dec-15
Dec-19
Nov-23
problems,
• appropriate safety mechanisms and procedures,
• use of existing field infrastructure & compromise to
Tim e (m onth-year) Linear Front enhance project economics.
Distance
A roap map for the implementation of an air injection project
Figure 7: Simulation Results: Linear Front Distance from Injector
follows in Figure 8. Some key considerations which apply to
vs. Time
different stages of implementation are highlighted, as well as
Important observations from the simulation results are, how these align with project phases.
• CO2 and N2 i.e. flue gas break through at producer
between 11-13 years (01/01/2019: CO2),
(01/01/2021: N2),
• no breakthrough of air i.e. no O2 at producer after 15
years (01/01/2023),
• combustion front location can be delineated by
temperature profile and confirmed by O2 and CO2
concentration profiles.
10 SPE 101099
P SCREENING/EVALUATION
R
O
J
E
C
T DESIGN
P
H
A
S IMPLEMENTATION
E
K
Experimental Health, Safety & Project
E
Testing Environment Monitoring
Y
C
O
N
S Compression Operational
I Requirements Issues
D
E
R
A NGL Recovery
T
I
O
N
S
R
O
A High Level Experimental Full Field Pilot Full Field
Screening Testing Implementation
D
M
A
P
Simulation
CURRENT
POSITION
Experimental Testing
The reaction behaviour of a crude oil is dependent on a series the only rigorous way to screen a candidate reservoir, but tests
of factors. The complexity of this system means that the also provide key data which allow screening for design e.g.,
reaction behaviour of a crude oil is reservoir specific, and may oil burning characteristics, and performance evaluation e.g.,
only be evaluated by testing with reservoir rock and fluid Air-Oil-Ratio (AOR). Data obtained are also necessary for
under reservoir conditions. A range of experimental testing accurate simulation e.g. reaction kinetic parameters.
techniques exist. These are designed to determine if and how Different types of tests are available and these operate at
the crude oil will react in the reservoir, and the behaviour of different pressures, temperatures, and hence can simulate
the system as any subsequent reactions occur. different conditions and provide different information. There
Experimental testing is the only way to determine whether is some variation between techniques, and terminology due to
an air injection candidate is suitable, and is integral to any air development by individual companies. A recommended
injection project. Not only does experimental testing provide minimum suite of tests for screening and design of an air
SPE 101099 11
APPENDIX A
Average Permeability
10 5 10-20 300 60 Range
(mD)
(10-2000)
Average Initial Water 0.37 (B), Current
0.45 0.35 0.19
Saturation 0.48 (C) (0.53)
API Gravity
30 39 32 30 31
(˚API)
Bubble Point Pressure
300 2246 625
(psia)
Solution GOR
120 525 205
(scf/stb)
Formation Volume Factor
1.16 1.4 1.205
(bbl/stb)
Water Water Water
Previous EOR None None None Water Injection
Injection Injection Injection
*Lsst = Carbonate (Limestone) Reservoir, Sst = Clastic
(Sandstone) Reservoir
APPENDIX B
Basin Surat-Bowen
Sand Precipice
Field Moonie
Depth 1800
(Metres)
Thickness (Metres) 20-60
Formation Type Sandstone
Permeability: 01-1 mD
Porosity: 7-11%
Initial Reservoir Pressure: 4200 psi
Current Pressure: 2700-3150 psi
Reservoir Temperature: 279 ºF
API Gravity: 50 ºAPI
Viscosity (@Pb and Tres): 0.139 cP
MW C7+ = 149.6
0
LEGEND: GAS SATURATIONS, FRACTION (N2, O2, CO2)
0
Y/X: 1.00:1
Axis Units: ft
1.0
-1,000
1.00
0.90
-1,000
0.80
0.70
0.9
-2,000
PROD01 INJ01 0.60
-2,000
0.50
0.40
0.8
-3,000
0.30
0.20
-3,000
0.10
0.00 195.00 390.00 yards
0.7
0.00 175.00 350.00 m eters 0.00
0.5
0 1,000 2,000 3,000 4,000
Date: 2005-09-15
Scale: 1:9007
0
Y/X: 1.00:1
Axis Units : ft
0.4
-1,000
1.00
0.90
-1,000
0.80
0.70
0.3
-2,000
PROD01 INJ01 0.60
-2,000
0.50
0.40
0.2
-3,000
0.30
0.20
-3,000
0.10
0.00 195.00 390.00 yards
0.1
0.00 175.00 350.00 m eters 0.00
0
Y/X: 1.00:1
Axis Units: ft
RES. A Base Case-Air Inj. Cooper Basin,
Gas Mole Fraction(N2) 2009-02-01 K layer: 1
File: Bas e Cas e.irf
User: bhughe01
0 1,000 2,000 3,000 4,000
Date: 2005-09-15
-1,000
Scale: 1:9007
1.00
0
Y/X: 1.00:1
Axis Units: ft 0.90
-1,000
0.80
-1,000
1.00 0.70
-2,000
-2,000
0.80 0.50
0.70 0.40
-2,000
-3,000
0.50 0.20
-3,000
0.40 0.10
0.00 195.00 390.00 yards
-3,000
0.10
0.00
0.00
195.00
175.00
390.00 yards
Y/X: 1.00:1
Axis Units: ft
RES. A Base Case-Air Inj. Cooper Basin,
Gas Mole Fraction(N2) 2023-01-01 K layer: 1
File: Bas e Cas e.irf
User: bhughe01
0 1,000 2,000 3,000 4,000
Date: 2005-09-15
-1,000
1.00
Scale: 1:9007
0
Y/X: 1.00:1
Axis Units: ft 0.90
-1,000
0.80
-1,000
1.00 0.70
-2,000
-2,000
0.80 0.50
0.70 0.40
-2,000
-3,000
0.50 0.20
-3,000
0.40 0.10
0.00 195.00 390.00 yards
0.30
0.00
0.00
195.00
175.00
390.00 yards
350.00 m eters
0.10
0.00
Figure D-6: CO2 Concentration, End of Injection
0 1,000 2,000 3,000 4,000