Professional Documents
Culture Documents
Recovery in a Heterogeneous
Argentina Waterflood
Presented by:
TIORCO, Inc.
The Improved Oil Recovery Company
INTRODUCTION
The Loma Alta Sur field is located in the province of Mendoza in the
Neuqun Basin of Argentina (Figure 1). The productive reservoir is
the Grupo Neuquen Formation, which is characterized as
heterogeneous multi-layer sandstone. In an effort to control the
vertical distribution of injected water, injection wells are completed
with downhole selective injection mandrels. However, the combined
effects of heterogeneity within the individual layers and the
extremely adverse mobility ratio motivated the operator to evaluate
techniques for in-depth volumetric sweep improvement.
Polymer is a traditional alternative for viscosifying water and
lowering the mobility ratio. However, relatively homogeneous
reservoirs are preferred in order to avoid polymer breakthrough in
offset producing wells. The objective of the operator was to apply a
staged chemical injection program that would reduce water
channeling in the highest permeability layers and, as a secondary
benefit, improve the oil-water mobility ratio.
FIGURE 1
FIGURE 3
The Loma Alta Sur structure is a north-northeast to south-southwest anticline approximately 3.5 km in length by 1 km
in width with dipping flanks of 20 to 25 where the presence of inverse faults has been detected (Figure 4).
Width varies from 150 m to 900 m in the most developed and continuous channel deposits. Generally, the
connectivity and areal development of the upper sand sequences is better defined than the lower layers. The fluid
distributions and drainage systems within the facies are influenced by structural as well as stratigraphic control.
Primary porosity is due primarily to calcareous cement dissolution, which is responsible for the abundant amount of
fines found in the pore throats.
TABLE 1
Reservoir Characteristics
Reservoir Type
Area, acres
Reservoir Temp, C
Average Depth, m
Total Pay Thickness, m
Permeability Range, md
Dykstra-Parsons Coefficient, v
Porosity, %
OOIP, m3
Sandstone
57.3
45
450
35
20 1000
0.89
18.8
2,169,000
Oil Characteristics
API Gravity
Viscosity, cp
21
30
FIGURE 4
CDGs were implemented in two phases in the LAS-58 pattern. In Phase I, 186,000 barrels of CDG were injected
during the period July 14, 2005--February 2, 2006. Phase II included 192,200 barrels of CDG from April 11--October
31, 2007. These CDG volumes represent 1.45% pore volume and 1.50% pore volume during Phase I and Phase II,
repectively. The total CDG volume injected in Phases I and II, therefore, was 2.96% of the LAS-58 pattern pore
volume. This is a very modest volume by industry standards (Braun and DeBons 1995; Chang et. al. 2004a).
DIAGNOSTICS
Injection Profiles
Repeated injection surveys demonstrate the inefficient
distribution of injected water in the LAS-58 injector. A
typical vertical profile is shown in Figure 5. In this
example, the three most permeable layers are receiving
approximately 90% of the injected water and several
layers are not being contacted by water.
Correlation of hydrocarbon intervals
The operator conducted a series of geological, reservoir
engineering and tracer surveys to correlate the Grupo
Neuqun Formation layers within the LAS-58 pattern.
Those studies confirmed that there was excellent
connectivity between the LAS-58 injection well and the
associated first line producing wells (Figure 6a and 6b).
FIGURE 6a
FIGURE 5
FIGURE 6b
Quantification of Heterogeneity
Core analysis of producing well LAS-72 (Figure 7)
indicated a Dykstra-Parsons coefficient (Dykstra and
Parsons 1950) of 0.89. A single core study is not
necessarily representative of reservoir heterogeneity;
however, the results of the LAS-72 core evaluation were
consistent with the injection profiles and oil/water
production behavior in the LAS-58 pattern. The operator
estimates that after three years of water injection the
LAS-58 pattern, cumulative secondary oil recovery is
only 5.48% of original oil in place (OOIP).
FIGURE 7
Tracers
Tracer velocity is proportional to the permeability and thickness of the reservoir flow paths. The permeability controls
the time of flight of the tracer particles between the injector-producer and the permeability-thickness product
controls the mass flow rate of the tracer (assuming a constant pressure drop). Interwell tracer tests can be utilized in
quantifying water channeling through the analysis of two tracer variables:
1. Times of flight distribution (or the first moment of the distribution)
2. Tracer mass produced (rate and cumulative production)
In addition to quantifying pre-CDG water channeling in the LAS-58 pattern, tracers were utilized to analyze the post
CDG times of flight and the tracer mass recovered. Due to the possible differences in the flow conditions between
the pre-CDG and post CDG tests, the tracer results are expressed in terms of the fractional daily tracer recovery (Fdtr)
and the fractional accumulated tracer recovered (Fatr). These variables are defined as:
Fdtr =
1
mtracer inj
Fatr =
1
mtracer inj
tracer sample
. q water producer dt
where:
mtracer-inj = tracer mass injected
Ctracer-sample = measured concentration in a producer
qwater-producer = water flow rate of the producer
T = time
mtracer-rec = tracer mass recovered in the time
A second tracer campaign was begun on June 14, 2006, approximately 4 months after the Phase I CDG (July 14,
2005February 2, 2006). However, in this test, different tracers were injected in each of the three mandrels of injector
LAS-58: Ammonium thiocyanate in mandrel 1 (lower layers), tritiated water in mandrel 2 (middle layers) and Yellow
Acid 73 in mandrel 3 (upper layers). Relatively rapid tracer breakthrough was observed in only two offset producers:
LAS-18 and LAS-49. The tracer response in each of these wells was similar, so only the results for producing well
LAS-49 are presented in Figure 10 and Figure 11, which indicate the fractional recovery of each of the three tracers by
mandrel.
A review of the post-CDG oil response for the LAS-49 (see CDG Results) does not corroborate the
apparent post-CDG water channeling implied by Figures 10 and 11. Several possible explanations exist. First, CDG
injection was restarted after this tracer study (CDG Phase II, April 11October 31, 2007). Tracers subsequent to
October 2007 may show that very different results from Figures 10 and 11. Secondly, the thief zones indicated by the
post Phase I CDG are believed to be direct, small volume channels. The fluid production from a small volume channel
in a single layer would be insignificant as a percentage of the total well production. Finally, the well production data is
based on periodic well production, which tends to mask production fluctuations. It should be noted that the total tracer
recovery in Figures 10 and 11 was 4.6% of the total tracer volume injected in all three mandrels, compared to 4% total
recovery of the pre-CDG tracer. The operator speculates that in the absence of the Phase I CDG pilot, the 2006 tracers
would have indicated much more pronounced water channeling.
FIGURE 10. LAS-49: Post-Phase I CDG. Fractional daily tracer recovery (Fdtr),
by mandril.
core floods was consistent with the reported field results (Smith 2000). CDGs have compared favorably to
uncrosslinked polymer in the Daqing field (Chang 2006). However, the CDG technology continues to be critically
evaluated in isolated instances (Seright 2006).
Colloidal dispersion gels (CDGs) are so named from the nature of the gel
solutions, which are suspensions of individual bundles of crosslinked polymer
molecules, or colloids. A solution of separate gel bundles are formed in
which a mixture of predominantly intramolecular and minimal intermolecular
crosslinking reactions.
By contrast, in bulk gels the crosslinking
mechanisms form a continuous intermolecular network of polymer molecules.
Figure 12 illustrates the difference in types of crosslinking reactions.
FIGURE 12.
Zone isolation
Most hydrocarbon reservoirs are composed of a series of layers, with each layer representing a distinct period of
geologic time with a unique depositional environment and type of minerals. In fact, stratification exists even when
there is no direct evidence of vertical separation. That is, even individual layers almost always include flowpaths of
varying porosity and permeability due to microscopic permeability distribution. Consequently, low volumetric sweep
efficiency occurs due to the petrophysical characteristics between different layers as well as permeability anomalies
within each layer. In Argentina, selective injection mandrels are frequently used in order to distribute injected water
more uniformly between layers. If the layers were homogeneous, this technique would result in high vertical sweep
efficiency, particularly in reservoirs with favorable mobility ratios. However, most reservoirs exhibit large
permeability variations even within individual layers. Seright (1988) asserts that zone isolation is necessary for gel
treatments in unfractured reservoirs that exhibit radial flow. In practice almost all gel treatments have been bullheaded,
indicating that (1) virtually all reservoirs are fractured, or contain high permeability anomalies with fracture-like
behavior, or; (2) significant permeability heterogeneity exists within individual layers. The authors believe that both
conditions frequently exist in waterfloods. Recent field experience (Romero 2003; Norman 2006) where gels
bullheaded in reservoirs with no known natural fractures indicates that gelant solutions behave like water during
injection, with preferential injectivity in the highest permeability layers. Sydansk (2006, 2007) postulates that most, if
not all reservoirs characterized as matrix rock are in fact fractured due to injection above fracture pressure, undetected
micro-fractures and/or extremely high permeability anomalies that exhibit fracture like behavior. Such reservoir
heterogeneities, difficult to quantify at a macroscopic level, offer a credible hypothesis for the mechanism of colloidal
dispersion gel propagation in the reservoir.
Reservoir engineering principles introduced in the mid 20th century demonstrate that, in the absence of free gas and
natural fractures, the fraction of total water injection entering each zone will be proportional to the flow capacity of that
zone (kh, the permeability thickness product). These mathematical models have been proven empirically in hundreds
of waterfloods.
Finally, studies have shown that polymer molecules cannot physically enter low permeability rock (Zaitoun, A. and
Kohler N. 1987). Due to the addition of a crosslinking agent a gel molecule is significantly larger than a polymer
molecule. Gel injectivity includes several variables, including polymer molecular weight, crosslinker effects and shear
rate, to name a few. Therefore, the lower permeability limit below which gel cannot be injected is reservoir specific
and difficult to quantify.
CDG RESULTS
Production Logs
Phase II of the CDG pilot was completed in October 2007. Therefore, the results presented in the following paragraphs
are attributable only to Phase I, which included 1.45% pore volume of CDG.
The operator ran a production logs before and after the
Phase I CDG treatment. Figure 13 indicates a positive
variation in the vertical profile. Injectivity was reduced
in the primarly thief zone (mandrel 3) and increased in
mandrels 1 and 2. Several layers (A170a, B120-B120a
and B150-B150a) had no injectivity prior to the CDG
pilot.
FIGURE 13
FIGURE 14
FIGURE 15
FIGURE 16
10
Figure 17
Figure 19
Figure 18
Figure 20
11
FIGURE 21
TABLE 2
2,169,000
21,194
62,000
2.9
CONCLUSIONS
1. Colloidal Dispersion Gels (CDGs) were applied in two Phases in a mature waterflood with an adverse
mobility ratio. Results to date indicate a clear oil response from the Phase I pilot. The Phase II pilot,
concluded in October 2007, is under evaluation.
2. No significant operational problems were encountered during the fourteen months of CDG injection (Phases I
and II).
3. Based on incremental oil quantified as of October 2007 from the Phase I pilot, the cost per incremental barrel
of oil is approximately $3.35. The ultimate cost per incremental barrel from the combined Phase I and Phase II
CDG pilots is expected to be in the range of $2.00 to $3.00.
4. The operator is currently performing an updated reservoir characterization in order to implement a field-wide
expansion of the CDG technology.
Avenida Mosconi 3169 Suite 5B Buenos Aires, Argentina 1419
54 11 4572 0027 54 11 4572 0015 Fax www.tiorco.com
12
REFERENCES
Braun, R.W. and DeBons, F.E., Polymer Flooding: Still A Viable IOR Tecnique, 8th European IOR Symposium,
Vienna, Austria, May 15-17, 1995.
Broseta, et. al., Shear Effects on Polyacrylamide/Chromium (III) Acetate Gelation, SPE Reservoir Evaluation &
Engineering, June 2000.
Chang, H.L, et. al, Successful Field Pilot of In-Depth Colloidal Dispersion Gel (CDG) Technology in Daqing Oil
Field, SPE 89460, 2004.
Chang, H.L., et. al., Advances in Polymer Flooding and Alkaline/Surfactant/Polymer Processes as Developed and
Applied in the Peoples Republic of China Journal of Petroleum Technology, Jan. 2006, pp. 84-89.
Chang, H.L., et. al., Successful Field Pilot of In-Depth Colloidal Disperion Gel (CDG) Technology in Daqing Oil
Field SPE 89460, 2004.
Dykstra, H., and Parsons, R.L., The Prediction of Waterflood Performance with Variation in Permeability Profile,
Prod. Monthly (1950)
Fielding, R.C., et. al., In-Depth Fluid Diversion Using an Evolution of Colloidal Disperion Gels and New Bulk Gels:
An Operational Case History of North Rainbow Ranch Unit, SPE 27773, 1994.
Mack, J.C., et. al., In-Depth Colloidal Dispersion Gels Improve Oil Recovery Efficiency SPE 27780, 1994.
Needham, R.B., et. al., Control of Water Mobility Using Polymers and Multivalent Ions, SPE 4747, 1974.
Nicol, A.B., et. al., The Adon Road-An In-Depth Gel Case History SPE 35352, 1996.
Norman, C. et. al., A Review of Over 100 Polymer Gel Injection Well Confomance Treatments in Argentina and
Venezuela: Design, Field Implementation and Evaluation, SPE 101781 (2006)
Romero, C. et. al., Non-Selective Placement of a Polymer Gel Treatment to Improve Water Injection Profile and
Sweep Efficiency in the Lagomar Field, Venezuela, SPE 80201 (2003)
Seright, R.S., Placement of Gels to Modify Injection Profiles, SPE 17332, 1988.
Seright, R.S., Are Colloidal Dispersion Gels Really a Viable Technology?, http://baervan.nmt.edu/randy/ March
2006.
Smith, J.E., et. al., Laboratory Studies of In-Depth Colloidal Disperion Gel (CDG) Technology for Daqing Oil Field
SPE 62610, 2000.
Smith, J.E., Closing the Lab-Field Gap: A Look at Near Wellbore Flow Regimes and Performance of 57 Field
Projects, SPE 27774, 1994.
Sydansk, R.D., et. al., More Than 12 Years of Experience with a Successful Conformance-Control Polymer Gel
Technology, SPE 59315, 1998
Sydansk, R.D., Key & Controversial Issues Relating to Conformance Improvement, Society of Petroleum Engineers
Advanced Technology Workshop, Comodoro Rivadavia, Argentina, November 2006.
Sydansk, R.D., New Conceptual Mechanisms for CDGs, unpublished presentation, Denver, Colorado, July, 2007.
Uliana, M.A., and Legaretta, L., 1993. Hydrocarbons Habitat In A Triassic-To-Cretaceous Sub-Andean Setting:
Neuqun Basin, Argentina, Journal of Petroleum Geology, October 1993
Veiga, R. and Orchuela, I.A., 1989. Identificacin de niveles generadores de hidrocarburos a partir de tcnicas de
perfilaje en la Formacin Vaca Muerta, Primer Congreso Nacional Exploracin Hidrocarburos. 2, 1061-1093.
Buenos Aires.
Zaitoun, A., Kohler, N., The Role of Adsorption in Polymer Propagation Through Reservoir Rocks, SPE 16274,
1987.
Zornes, D.L., et. al., An Overview and Evaluation of the North Burbank Unit Block A Polymer Flood Project, Osage
County, Oklahoma SPE 14113, 1986
ACKNOWLEDGEMENTS:
The authors thank the management of Repsol-YPF for the opportunity to present this paper. Also, we would like to
acknowledge the assistance of the Malargue Area Operations and Development Groups as well as the consulting firm
NCT of Maracaibo, Venezuela.
13
NOMENCLATURE
CDG
= colloidal dispersion gel
Cp
= centipoises
Ev
= volumetric sweep efficiency
H
= thickness
k
= permeability
M
= mobility ratio
m
= meters
3
m
= cubic meters
Md
= millidarcies
Np
= cumulative oil production
CONVERSIONS
km2
1.00 E + 06 = m
OOIP
ppm
RRF
Sor
Qinj
Qo
Qw
V
VP
WOR
=
=
=
=
=
=
=
=
=
=
km
1.00 E + 03 = m