Sei sulla pagina 1di 7

SPE 165462

A Heuristic Production Model for Cyclic Steam Stimulation in a Fractured


Heavy Oil Carbonate Reservoir
Kent Qin and Jeff MacDonald, Osum Oil Sands Corp

Copyright 2013, Society of Petroleum Engineers


This paper was prepared for presentation at the SPE Heavy Oil Conference Canada held in Calgary, Alberta, Canada, 1113 June 2013.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
The Upper Devonian Grosmont formation is considered to be Canadas next largest unconventional oil resource play, with an
estimated 406 billion barrels of heavy oil in place. A number of production pilots targeting the Grosmont formation have
tested various thermal enhanced oil recovery techniques, which include steam flood, combustion, and cyclic steam
stimulation (CSS). To date, the CSS process has demonstrated considerable promise based on performance from the Unocal
Buffalo Creek Phase 2 pilot and through the application of C-SAGD, a CSS variant designed for the Grosmont, at the
Laricina-Osum joint-venture pilot in Saleski. While the methodologies for forecasting CSS production profiles are well
understood for clastic oil sands reservoirs, no direct analogue exists for carbonate reservoirs such as the Grosmont. The
current profiling model for clastics appears to be suitable for matching and forecasting the production characteristics of CSS
and the early cycles of C-SAGD in the Grosmont. However, a few extensions to the current profiling models are required to
address differences in cycle length, bitumen ramp up, and oil cut characteristics. For instance, oil cuts are typically low
initially and increase with time for the majority of cycles in clastic oil sands, whereas in the Grosmont, there is observed
cycle-to-cycle variability. The first few cycles typically show high initial oil cuts, decreasing with time; in subsequent cycles,
the oil cuts behave similarly to clastic reservoirs. These differences can be attributed to the presence of secondary porosity
and permeability systems in the carbonate formation, such as vugs and fracture networks, and their interactions with the
matrix. This paper will describe the modification to existing models for profiling production, based on field observations
from the aforementioned pilots.

Introduction
With over 400 billion barrels of estimated oil in place, a variety of thermal pilots have been implemented since the 1970s to
assess the Grosmont Formation (Yuan 2010). The efforts to commercialize this resource have been revived recently with
initial success from a thermal pilot in Saleski, operated by Laricina Energy Ltd. in partnership with Osum Oil Sands Corp.
However, with limited field history and the complex nature of carbonates, profiling methods cannot confidently be generated
based on pilot results alone. A method is proposed to incorporate experience from decades of commercial CSS production
from the Cold Lake area, where existing production exceeds 300,000 bbl/d. This paper attempts to build upon this extensive
knowledge base by taking a heuristic view of past and current pilot history in the Grosmont C (GMC) Formation. In
particular, key mechanistic differences to Cold Lake performance are provided, in an effort to facilitate a forecast
methodology.

Current Clastic CSS Forecast Model


The conventional approach for forecasting CSS field performance in clastic oil sands projects takes into account both the
short term (intracycle) and long term (intercycle) behaviours. The intercycle behaviour refers to the cycle-to-cycle variability
in oil recovery and steam-oil ratio (SOR). Analysis of long-term CSS field performance in the Cold Lake area has shown that
overall intercycle behaviours exhibit a repeatable pattern in spite of variations in local reservoir properties, such as net pay,
bitumen saturation, and geological heterogeneities. Consequently, it is possible to describe intercycle behaviours using a
generalized recovery type curve as shown in Fig. 1. Oil recovery factor (RF) is plotted against pore-volume-fraction steam
(PVSteam), defined as cumulative steam injected normalized against drainage box pore volume. The slope of the profile is

SPE 165462

analogous to oil-steam-ratio. This normalization is required to properly compare performance between projects of different
reservoir characteristics, well type (vertical or horizontal), and drainage box dimensions. Scaling factors for RF are often
used to match the type curve with actual and/or analogue production history to reflect the well/pad performance of interest.
The intracycle behaviour refers to the total fluid flow rates and oil cuts as a function of time within each CSS cycle. Fig. 3
shows series of normalized rate-cumulative plots for typical CSS cycles in the Cold Lake area. The rows represent different
wells, and columns represent successive cycles. The cumulative total fluid volume has been normalized against total steam
injected for each cycle, defined as the total-fluids-to-steam ratio (TFSR). The black lines show the trends in total fluid rate
decline, whereas the green lines show the trend in oil cuts.
In typical CSS operations in the Cold Lake area, dilation and/or fracturing of the reservoir occurs during high pressure steam
injection. The production phase of a CSS cycle begins with high rate of fluid evacuation from the fracture, driven by
overburden stress on the fracture face. Initial flow rates for CSS are typically restricted due to surface equipment or artificial
lift constraints. This non-boundary dominated flow cannot be reflected by decline analysis, and is represented by the flat line
regions on the rate-cumulative plots. Once a well has surpassed the flowing-pumping stage of production, the flow is
boundary dominated in the reservoir and the total fluid rates can be used for decline analysis. A linear decay in the ratecumulative plot suggests that an exponential decline model is applicable for forecasting total fluid with time. In a forecast
model, the total fluid behaviour is separated into two intervals. A constant flow rate interval used for flow back and flowingpumping, and a constant decline rate interval after the flow is boundary dominated in the reservoir. The transition between
the two intervals is a function of operating conditions, artificial lift capabilities and hence is determined from examination of
production data.
Vittoratos (1991) and Batycky et al. (1997) proposed a conceptual and mechanistic model to describe the intracycle behaviour
for oil cuts in CSS. A brief summary with supporting production data is provided here. For the majority of a wells life, the
initial fluid recovery in each cycle is characterized by the production of free water from the near wellbore and depleted zones,
and from the evacuation of steam condensate from fractures. The production of the free water is followed by a period of
erratic free water and bitumen slug production. These two production intervals are reflected in low and erratic oil cuts. An
exception to this trend is observed on occasion in the first cycle only, where a brief interval of high oil cut can exist in the
initial days of production (not resolvable from monthly averages). After the production of the free water, a ramp up and
plateau of oil cut are observed. The fluid regime during the plateau is entirely emulsion. The extended and controlled release
of solution gas from bitumen at the edge of the heated zone provides the driving force for the bitumen into the wellbore (see
Fig. 2). The continuous condensation of injected steam at the edge of the heated zone is a major contributor to the uniform
production of water and hence oil cuts during this period. In an empirical forecast model, the intracycle oil cut trend is
parameterized into three linear segments with respect to TFSR: the low oil cut, the ramp up, and the plateau oil cut.
Numerical values for the low and high oil cuts, and the timing to ramp up are determined from fitting against actual data. The
rate of ramp up is a function of cycle SOR and cycle TFSR, and is coupled with intercycle forecast model, and the intracycle
forecast model for total fluid decline.
Observed Production Performance in the Grosmont Formation
Production performance from the two Grosmont carbonate pilots will be assessed for applicability of the existing empirical
model for clastic CSS.
Unocal Buffalo Creek Pilot
The Buffalo Creek Phase 2 Pilot (10-05-088-19W4), operated 10 CSS cycles from 1980 to 1985 by Unocal in partnership
with Canadian Superior and AOSTRA. The pilot was conducted using a single vertical well for injection and production, and
was perforated over a 10 m interval within the GMC Formation. Table 1 shows the production summary for the Buffalo
Creek pilot. The injection and production volumes for cycle 6 have been excluded from further analyses, as hot water was
injected during this cycle, which is not representative of the CSS process.
Fig. 1 shows the RF-PVSteam plot for the Buffalo Creek pilot, compared with a matched type curve derived from CSS
operations in clastic reservoirs. A typical profile for horizontal well CSS based on 160m well spacing is also shown. The
match indicates that intercycle behaviours are in agreement with the trends from clastic reservoirs. Note that in order to
forecast full field, the effect of well confinement should be considered. For example, Imperial Oil Ltd. (2012) has
demonstrated that confinement can result in an additional 50% uplift in recovery factor for the same volume of steam
injected. A similar uplift should be expected in the Grosmont. Fig. 4 shows the rate-cumulative and oil cut-cumulative plots
for the Buffalo Creek pilot. The black lines indicate the trends in total fluid rate decline, whereas the green lines indicate the
trend in oil cuts. Daily data were only available for cycles 1 to 5 (Union Oil 1981a, 1981b, 1982); monthly data from the
public database were used for the remaining cycles (7 to 10). The rate-cumulative plots for Buffalo Creek all showed linear
decline total fluid rates with respect to cumulative fluid production. This supports the exponential decline model for total

SPE 165462

fluid decline with time and is in agreement with observations from CSS in clastic reservoirs. Furthermore, Fig. 5 shows that
the decline rate is equivalent for cycles 2 to 5 for similar steam injection volumes. In terms of oil cuts, for early cycles (1 to
3), the oil cuts at the beginning of each cycle were high (up to 75%), followed by a gradual decline in oil cuts with respect to
TFSR. This initial high oil cut with gradual decline trend can be observed for multiple cycles. This differs from the
behaviours of CSS in clastic reservoirs, where the high initial oil cuts are seldom or of short duration in the first cycle only.
From cycle 5 onwards, the intracycle oil cut behaviour from Buffalo Creek began to exhibit similar characteristics to clastic
reservoirs, where the oil cuts were low in the beginning of production and ramping up to a plateau as the cycle progressed.

Saleski Pilot
Laricina Energy Ltd., in partnership with Osum, is currently operating a production field pilot in Saleski (section 26-08519W4M) targeting the Grosmont C and D Formations. The most recent well, P2C, is a horizontal well in the GMC. It has
been operated at below fracture pressures using a C-SAGD process, which is a CSS variant designed for the Grosmont. Fig. 6
shows the oil cut-cumulative plots for the three production cycles from the P2C well. Early cycle oil cut trends show
agreement with those from Buffalo Creek: high initial oil cuts (up to 90%) were observed in all three cycles, and gradually
decreasing over time. Intracycle oil cuts have not transitioned to the long-term behaviour seen in Buffalo Creek and clastic
CSS.

Carbonates: Modifications Required to Clastic CSS Forecast Model


The examination of data from CSS in carbonate reservoirs above has highlighted a number of similarities and differences in
the intercycle and intracycle behaviours as compared to those seen in clastic reservoirs. The close fit between the Buffalo
Creek production data and the clastic CSS type curve on a RF-PVSteam plot (Fig. 1), suggests that the approach for
forecasting long-term intercycle performance for CSS in clastic reservoir should be applicable to the GMC reservoir.
Likewise, both Buffalo Creek and clastic performance exhibit similar linear decline response to TFSR (exponential with
time). Accordingly, the intracycle total fluid decline trends for the GMC can be assumed to be comparable to clastic CSS.
High initial oil cuts at the start of production are common to both the Buffalo Creek and Saleski pilots, indicating a unique
yet repeatable response from the GMC Formation. In the case of the carbonates, the initial oil cuts are much higher than
observed in first cycle in clastic reservoirs; its decay is more gradual, and its presence is repeated across numerous cycles.
The high-to-low trend for intracycle oil cuts eventually transition to low-to-high with increasing cycles. The following
discussion attempts to describe a conceptual model to understand this behaviour.


A Modified Conceptual Model for the GMC
Fluid flow within carbonate reservoirs can vary significantly from that in clastic reservoirs due to heterogeneous
characteristics (Lee, 2010). Fig. 7 shows a core photo from the GMC Formation, providing a visual representation for the
presence and prevalence of secondary porosity and permeability systems. The rock matrix has a measured air permeability on
the order of 100 mD. However, the effective permeabilities of the system have been determined to be much higher, on the
order of 100 D, based on Shells water test in the GMD Formation (Machel 2009), Laricinas history match of Buffalo Creek
(Novak et al. 2007), and Osums recent water interference test in the GMC Formation. This highlights the large contribution
to the overall permeability by the vugs and the interconnected fracture networks. It is due to these high permeabilities that
fracturing the reservoir with high-pressure CSS was not required to inject desired steam volumes.

When steam is injected into the GMC Formation, the significant permeability contrast between the matrix and secondary
porosity ensures that steam preferentially flows into the latter. Since it is reasonable to assume that the void spaces are
bitumen-saturated (Russel-Houston et al. 2009), the high viscosity ratio between steam and bitumen promotes the fingering
of injected steam through the bitumen. The oil-wet nature of the carbonate would further promote steam fingering through
the bitumen rather than along the fracture face (Stokes 1986). The latent heat from the injected steam is transferred by
convection to the bypassed bitumen around the steam fingers, and by conduction to the matrix and vugs perpendicular to its
path. According to Vittoratos (1991), the distinction between the heating mechanisms (conduction vs. convection),
determines the extent of water entrainment in the emulsion. Convection is thought to cause greater water entrainment by
turbulence with condensed water, hence low emulsion oil cuts, and the opposite is true for conduction.
The high permeability of the fracture network in the GMC results in the injected steam condensing further into the reservoir.
This is analogous to the phenomenon within steam-induced fractures in high-pressure CSS in clastic reservoirs. In both
scenarios, the initial bitumen influx is heated via conduction, resulting in the produced emulsion having high oil cuts. In
contrast to a non-repeatable oil spike typically observed in clastic reservoirs (first cycle only), high initial oil-cuts with
gradual decline across multiple cycles were observed in the Buffalo Creek and Saleski pilots. This is supported by the fact
that vugs in the GMC, having high storativity and permeability, and therefore would provide a more effective supply of
bitumen into the fractures than a low permeability matrix in a conduction-only process. Furthermore, the steam-induced

SPE 165462

fractures in the clastic reservoirs heal as pressure declines (hence the oil spike), whereas the fracture networks in the GMC
are permanent, resulting in a more gradual decline.
As the vugs become de-saturated with increasing cycles, a large free-water zone can exist in the near-wellbore region poststeaming. Oil production is deferred further until this free-water zone is produced, resulting in low-to-high oil cuts behaviour,
similar to that observed in clastic CSS reservoirs. Although injected steam will still contact bitumen within secondary
porosity, the low-to-high oil cuts possibly coincides with decreasing bitumen contribution from the vugs, and increasing
bitumen contribution from the matrix as convection becomes the major heat transfer mechanism. Along with increasing net
heat input into the formation, the matrix attains a temperature conducive for longer production cycle length. This
phenomenon was observable for cycle 5 and onwards from the Buffalo Creek pilot.

Implications to Reservoir Management in the GMC


As previously described, conduction is the dominant heat transfer mechanism for formation heating during the early cycles.
Given the steam slug sizes from the Buffalo Creek pilot, heating by conduction alone was not sufficient to deliver the
required thermal energy into the formation to maintain emulsion mobility for an extended period of time. This was evident in
the low TFSR observed in the initial three cycles from Buffalo Creek. The strategy to extending the cycle length entails
larger steam injection volumes. Since the GMC Formation is extensively fractured, the reservoir could easily accommodate
larger steam volumes relative to clastic reservoirs.
Furthermore, because the emulsion produced during the initial cycles had much lower water cut, its bulk viscosity was
dominated by that of bitumen, which is a strong function of temperature. The cooling bitumen in the fractures eventually
forms a plug, effectively trapping steam condensate and leads to curtailed production. This is represented in the Buffalo
Creek summary Table 1, where cycle water-to-steam ratios (WSR) do not exceed 0.3 for the initial three cycles. This has
large implications on the development of the reservoir, since significantly higher water make-up should be expected during
project start-up. The conventional CSS approaches for field-scale pressure management, such as parallel drilling (for
horizontal wells), megarow steaming strategy, and a pressure set-back buffer should improve water recovery performance in
the Grosmont.

Conclusion
The mechanisms identified in clastic reservoirs for flow regimes and oil drive can be used to explain the phenomena
observed in the GMC. The good agreement between the intercycle SOR behaviours and intracycle total fluid decline in both
clastics and GMC reservoirs supports the adaptation of the forecast models for use with the GMC. For the intracycle oil cut
behaviours, later cycle oil cut trends follow the low-to-high model in clastic reservoirs, whereas in early cycles, high-to-low
trends have been observed repeatedly in both the Buffalo Creek and Saleski field data. The forecast model should be
modified to accommodate repeatable high-to-low trends observed in earlier cycles.

Acknowledgements
The authors would like to thank Osum Oil Sands Corp. and Laricina Energy Ltd. for their permission to present this paper.

References
Batycky, J., 1997. An Assessment of In-Situ Oil Sands Recovery Process. Distinguished Author Series, J Can Pet Techno, 50 (9), 15-19.
SPE-97-010-DAS
Batycky, J., Leaute, R.P., and Dawe, B.A., 1997. A Mechanistic Model of Cyclic Steam Stimulation. Paper SPE 37550 presented at the
International Thermal Operations and Heavy Oil Symposium, in Bakersfield, California, 10-12 February.
Imperial Oil Limited, 2012. Cold Lake Approvals 8558 and 4510 Annual Performance Review.Lee, W.S., 2010. Numerical Simulation
Techniques in Heavy Oil Carbonate Reservoirs. Geosystem Engineering, 13 (2), 69-76.
Machel, H.G., Borrero, M.L., Dembicki, E., Huebscher, H., Ping, L., Zhao, Y., 2012. The Grosmont: the Worlds Largest Unconventional
Oil Reservoir Hosted in Carbonate Rocks. Geological Society, London, Special Publications (in press; published online 21 September
2009). http://dx.doi.org/10.0.4.120/SP370.11
Novak, J., Edmunds, N., and Cimolai, M., 2007. A History Match of CSS Recovery in the Grosmont. Paper CIPC 2007-154 presented at
the Petroleum Societys 8th Canadian International Petroleum Conference, Calgary, Alberta, 12-14 June.
Russel-Houston, J., Rott, C.M., Putnam, P., McGrory, R., 2009. The Grosmont C at Saleski, Northern Alberta: a Solution-Enhanced,
Highly Fractured Dolomite Reservoir. Canadian Society of Petroleum Geologists Annual Convention, Calgary, Canada. Abstract.
Stokes, J.P., Weitz, D.A.,Gollub, J.P., Dougherty, A., Robbins, M.O., Chaikin, P.M., Lindsay, H.M., 1986. Interfacial Stability of
Immiscible Displacement in a Porous Medium. Phys. Rev. Lett. 57 (14), 17181721

SPE 165462

Union Oil Company of Canada, 1981a. Buffalo Creek Test Site: Progress Report ER-81-17; ERCB Approval No. 2367C, AUB, EUB, AB
Union Oil Company of Canada, 1981b. Buffalo Creek Test Site: Progress Report ER-81-38; ERCB Approval No. 2367C, AUB, EUB, AB
Union Oil Company of Canada, 1982. Buffalo Creek Test Site: Progress Report ER-82-08; ERCB Approval No. 2367C, AUB, EUB, AB
Vittoratos, E.S., 1991. Flow Regimes During Cyclic Steam Stimulation at Cold Lake. J Can Pet Techno, 30 (1), 82-86. SPE- 91-01-07
Yuan, J-Y., Jiang, Q., Russel-Houston, J., Thornton, B., Putnam, P., 2010. Evolving Recovery Technologies Directed Towards Commercial
Development of the Grosmont Carbonate Reservoirs. Paper CSUG/SPE 137941 presented at the Canadian Unconventional Resources
and International Petroleum Conference in Calgary, Alberta, 19-21 October.

Tables

Table 1: Buffalo Creek Pilot production summary.


Cycle

Steam Inj.
(m3)

Bitumen Prd.
(m3)

Cycle
SOR

WSR(a)

TFSR(b)

1
2

6,990

489

14.3

0.09

0.18

7,048

1,331

5.3

0.2

0.41

7,073

1,224

5.8

0.28

0.47

7,099

1,945

3.7

0.56

0.85

7,104

1,610

4.4

0.68

0.92

6(c)(d)

9,480

779

12.2

0.58

0.66

(c)

7,169

2,015

3.6

0.86

1.14

(c)

7,314

1,499

4.9

0.98

1.18

(c)

13,943

2,573

5.4

0.69

0.87

14,000

2,053

6.8

0.84

0.99

10(c)
(a)
(b)
(c)
(d)

Water to steam ratio (m3/m3).


Total fluid to steam ratio (m3/m3).
Injection and production volumes estimated from Accumap monthly data.
Below average performance due to injection of hot water in place of steam.

Figures
0.35

TypicalTypeCurve:ColdLakeCSS
FieldData:BuffaloCreek
MatchedTypeCurve:BuffaloCreek

0.30

RecoveryFactor

0.25
0.20
0.15
0.10
0.05
0.00
0.0

0.2

0.4

0.6
0.8
PVSteam

1.0

1.2

1.4

Fig. 1: RF-PVSteam plot for a typical horizontal, 160m well spacing Cold Lake CSS well (green). Buffalo Creek vertical CSS pilot
performance (solid blue) can be fitted to existing clastic CSS type curve (dashed blue).

SPE 165462

Fig. 2: CSS mechanism conceptual model (Batycky 1997).

800

Well2

OilCut
TotalFluid

600

0.6
400
0.4
200

0.2
0

0.4

TFSR

0.6

0.8

600

0.6
400
0.4
200

0.2
0
0.2

0.4

TFSR

0.6

0.8

TotalFluid

600
400

0.4
200

0.2

0.4

0.6

TFSR

0.8

TotalFluid

600
400

0.4
200

0.2
0

0
0

0.2

0.4

TFSR

0.6

0.8

0.6

TFSR

0.8

600
400

0.4
200

0.2

0.4

0.6

TFSR

0.8

400
0.4
200

0.4

TFSR

0.6

0.8

400
0.4
200

0.4

0.6

TFSR

0.8

OilCut
TotalFluid

400
0.4
200

0.2
0

0
0

0.2

0.4

0.6

TFSR

0.8

0.0
0.1

0.15

0.0

0.2

0.1

0.2

50

0.2
0.0
0.4

0.5

0.6

100

50

0.2
0.0
0.2

0.4

150

OilCut

0.6
100
0.4
50

0.2
0.0

0
0.4

0.6

0.8
TFSR

1.0

1.2

1.4

TotalFluidRate(m3/d)
OilCut

TotalFluids

0.6

0.8

TotalFluidRate(m3/d)

OilCut

100

50

0.2
0.0

0
0.0

0.2

0.4

0.6

0.8

1.0

1.2

TFSR

150

0.6
100
0.4
50

0.2
0.0

0
0.6
TFSR

150

0.4

TotalFluids

0.4

200
OilCut
TotalFluids

1.0

Cycle10

OilCut

0.8

0.2

0.5

0.6

200

Cycle9

0.0

0.4

Cycle7

TFSR
1.0

0.3

0.8

0
0

OilCut

0.8

0.2

150

0.4

0.7

200

Cycle8

0.0

0.2

1.0

OilCut

TotalFluids

TFSR
1.0

0.1

TFSR

0.6

0
0.3

0
0

0.8

1.0

0.8

200
OilCut
TotalFluids

150

0.6

OilCut

OilCut

100
0.4

0.2

0.0

OilCut

0.8

OilCut

150

TotalFluidRate(m3/d)

TotalFluids

0.6

0.1

50

0.2

200

Cycle5

Cycle4

0.8

100

TFSR
1.0

200

Cycle4

150

0.4

0.4

TotalFluidRate(m3/d)

TFSR
1.0

0.3

TotalFluids

0.6

0
0

TotalFluidRate(m3/d)

0.05

50

0.2

0
0

100
0.4

Cycle3

TotalFluidRate(m3/d)

0.2

150

0.6

200

Cycle3

0.8

100
0.4
50

0.2
0.0

TotalFluidRate(m3/d)

50

1.0

TotalFluidRate(m3/d)

100
0.4

TotalFluids

0
0.0

0.2

600

0.6

OilCut

0.8

OilCut

150

TotalFluidRate(m3/d)

TotalFluids

200

Cycle2

OilCut

0.6

OilCut

1.0

200

Cycle1

0.8

1
800

Well3

Fig. 3: Several Cold Lake (horizontal) CSS wells highlighting the typical increasing oil cut behaviour as a function of TFSR.
1.0

0.4

TotalFluidRate(m3/d)

0
0.2

0.8

0
0.2

600

0.6

0.6

OilCut
TotalFluid

600

OilCut

0.2

0.8

800

Well2

TotalFluid

0.6

TFSR

0.2

800

Well3

0.4

0.8

0.8

Oil Cut

Oil Cut

0.2

OilCut

0.2

0
0

TotalFluid

200

0.2

800

Well2

OilCut

400
0.4

0.6

0.6

800

Well3

0.4

0.8

0.8

0
0

OilCut

0.2

600

0.6

0
0

1
800

Well2

OilCut

0.6

TotalFluid

0.8

800

Well3

0.8

0.6

TFSR

0.2

Oil Cut

0.2

0.4

0.8

0
0

0.2

200

0.2

0
0

0.4

TotalFluid

TotalFluidRate(m3/d)

400

OilCut

0.6

0.8

1.0

TFSR

Fig. 4: Buffalo Creek oil cuts as a function of TFSR for cycles 1 to 10, demonstrating cycle variability. Cycle 6 omitted as hot water
was injected. Early cycles exhibit high initial oil cuts and decrease over time. Later cycles show clastic-like behaviour with low to

TotalFluidRate(m3/d)

0.8

0.6

800

Well1

0.8

Oil Cut

0.6

600

TotalFluidRate(m3/d)

TFSR

Oil Cut

0.4

TotalFluidRate(m3/d)

0.2

0.8

Oil Cut

0.2

0
0

200

TotalFluid

Oil Cut

400
0.4

OilCut

TotalFluidRate(m3/d)

0.2

0.6

CYCLE#4
1

800

Well1

0.8

Oil Cut

200

600

TotalFluidRate(m3/d)

0.4

TotalFluid

Oil Cut

400

OilCut

TotalFluidRate(m3/d)

0.6

CYCLE#3
1

800

Well1

0.8

Oil Cut

600

TotalFluidRate(m3/d)

Oil Cut

TotalFluid

TotalFluidRate(m3/d)

OilCut

0.8

Oil Cut

TotalFluidRate(m3/d)

CYCLE#2
800

Well1

TotalFluidRate(m3/d)

CYCLE#1
1

SPE 165462

high oil cuts.

160
Cycle1

140

TotalFluidRate(m3/d)

120

Cycle2

Cycle3

Cycle4

Cycle5

Increasingcycles

100
80
60
40
20
0
0.0

0.1

0.2

0.3

0.4

0.5
TFSR

0.6

0.7

0.8

0.9

1.0

Fig. 5: Consistent linear trends for Buffalo Creek total fluid as a function of TFSR for cycles 1 to 5.

1.0

Cycle1

1.0

Cycle2

0.8

0.6

0.6

0.6

0.4

0.4

0.2

0.2

0.0

0.0
TFSR

OilCut

0.8

0.8

OilCut

OilCut

1.0

Cycle3

0.4
0.2
0.0

TFSR

TFSR

Fig. 6: Saleski GMC well oil cuts as a function of TFSR (removed for publication) for the first three production cycles. Early cycle
behaviour exhibiting high (up to 90%) to low oil cuts consistent with observations from Buffalo Creek.

Fig. 7: Grosmont C core demonstrating vuggy porosity and vertical fracturing.

Potrebbero piacerti anche